>. mee wes

aero

OR ee ee Pe Pre Or Omen

ae as ale ete A ee Bee et ee
ee ee ee " +
ne 6 = ee ee er Oe ot pete eer = Meters ete

Sedadantindhdndedhe ON me © om OW etere terete +
ee ee ee od

~ oasis ses mathanwoes tim ne meh

PF Le OT OOPS FORTY ew ee eee eee etre preteen re ee the

Ae nee VIP TOD TERT ay let Spe w~ atewrete wetwionnyy ern eer yr eres TOTO
Tn ee een Cw Sey ewe ete © OOS Oe tee “rere

erate ene Oe adadh ttt
wre OUT
ieee etree tatoo Pin OSE at
or rer naar NS tian
tere oe - co ow ore et ores ann te te tne
rewe - rer rwrrene* “_
wows oe oH eee ton POE PTT OR ONCE oe ee Pe eee ee TT OFS ONT ET OE EIT eer ©
aetetedeben VIN ee OPT ee Coren ete er ee

jecete teen neremate terete Sh head - omit “ vd iee
opie tepte ® wee ° . reine the eters -
= ee oisintereiereneterernrertia ater

der
ee

vos

sal etapa ep terorrr e+
Oe ew
. Aarereeee o

Re ae See ee ee
ere SNE OFe we ets Se

owiehe ‘ :
panperne Seana seein PS eheetrlene re et Of ee wr

‘
wee a

vere

‘ eT ledethdieal z ny she ~ aavetty a.
pareraeaner ea teenth ate Cn aetna tet tin tind vattinli ras tind ine are ens eh A ae TT adel
~~ nr ! vate qrenetgreenwre treet ee r rt ee nr ie Secret tens oe renmuurows : : ee OTTO Se HVS TETY TTT
. notion entire itt Lee eee eve Here Tika eh 7a eee ee ve wwererew mS» wo cheraragurgnerea: @ ©" RatOrewereerareoemneen gree ee OT TTT NOTT LOOT AON
eT cnasndhtnd>tecethd ree eeny ee eee b the aes oo . SIE omen eT OTe oe re ee ow at eT atntntvtatel
evurrnrre oor ee ee ows mS mt mae a ae oooroe™ Tew
: ene eee eatatintiadl ~
——_—

Sony y Rhee emwimL EMH ORT eeTY eee

vere rT eee © rere dlc ddetietbtedtadidntndadadetedadh tadeceteechdadatace A dahanina dda Leo) on Adnan 0m

, aati dian itt diadiadtadiadid ornerwe
pote + ote wlioeeoetwes neg oener erat WF aedibdid id te el ae ee ee orwre 2? pcan Ge pa ~
. wae ; ’ * pers on eo 1 ad ad
+e Pe ahem pn aoa verre POUT Ver Erte ore rn ergrerirenete _ Sa ey tra COMPETE Owe cote t wow owe stully- quindlb Gtatbneeppramrertechouairemanmermen > eee irae annie owe oe eee"
ets wee nee nn ants ga ah nea 7 were ey -- ~*. _ * 4 Sa eene mae meena sens Ah ee ee ee ee en alent tide beat ay Som nnn S ow Teor ewe ove
“or ew ee tts + fe HO Ow OF & nit dtiedindinnngedat ran eee oP " ad . “—~e “~ * . rere . ; —
ree Lain tebe rt ah - " ee ee be ee a 1 PPE OOF Ore Pe ewww oes a a ce oc
ooo weer oe

verre ee
er eo mee

reer grees
PONT TE! OF ee we
wee

. vo" ew ewe ee oe ee :
FOV CR SE ETP EET ECU MN eee PETES eT FTE VUE PU CUNT OTT U9 Te
Ee ee He ETOP W En ee ww ee ee NG eer owe ners oe ©
FOO CPOE CU He Oe wee CCT ROSS * oP UTE CWE ee TET OUND FU COVE

oa

Te eee iedicdastindinindiadih did indindiasinan tedi> tedatdinae
UT ERT OR Oe Pererene wisi renrte ee
ere ree" PROTETPT RY Ferre .
Pe ahd

paneirteer® rs relgretetewigde-qxuseteierere
ceyapnersiqipia bennsereneny” viel rele cree see = rome nee

+ ow Teher re ws
en se

v
" .
i

nerererarete we ©

b4 rts OE OUT UT vee ¢ .
Agha tetetatetet mos . povtere ete omy tw ote . eden eer ow erIpe SF FES OE HFS SOE! eee ere Ee HIT Hele we > whe a - ~- 7
coreneey tumetonatarecenepnehet htt? wr errors Oe? om we ee cee aM 8 Per + Oe Cee & ow eee Bw : > pelrneeneemesaeet = te tr Vepewe ren yee oer ayers ew
4 apfetePerplnry ene everErE ew Terr F . ~—e ee es ore mt Se eine tgp sittin ooewur ee ooer ewe oe
whereree? ms aaah oe “ ee ome’ Hyerees tr owe e eRe seve rine hy se ed ~~ ene acetate een te DOO OAS

‘ . : " - <n ot — on : - eee wee wr
ew Path shinpatorepnteabine sis er orerreter tetera ee re Oy > ?s nt ae eo a : ae verwrrr ras Pe YEes HEer rere rere

Corr ws rerwrmere Oe ae Ma eel EY FOF Cee e-ere etetene ee ome oe Per oe eres Hw rrr es eR YT Fer rr + errr ew “
- POETS FOO Sy OOS CHEE oor’ OOO ee eee ee

sewer wes

vnerwereneres qr ee ier

leer eee dnd _——e. eer ere eT ee Or Oe ere eee TTS F

oor UWE HE rye He ree eo reetaeneng eee eo eet Owe ower

Pew ees 19'S 0 OCU ST"
wewe oor

COP RCW Het Hit ee te ew He

PO OT HP Se wir eRe EE Or eee Hee «

oer
——eere + oe FR eee

rT est Te Ke

se indidainde dete tat

or

Peer she eee ete oe oye” rte
— oe Ep hE Ve CO eet er et Vee? Oe we em weer eet . :
or ethene vee oor _ " Dé TT OS Ore vy ee ° * - ’ .
. *. oo ~ , 7 A i , ° ue hited es oo ~~ —<-*

“ _httatndadtatad tetracaine ~ $0 eins wcinslacttrtndnd iddiptadnchththdiptinhGhthdindadietdndh dade: o.4 tac Pe ooh rien te Ri A a tnt
19 an : pete View ee Ter eter Petes talerrs mew wre . were errr: OU Ee UES PER ERT 60 TF PV metew nes - . spe £eNe ere erw'eeeuwe= —er
ernre ge ey ere Oho MESS 'e eHow Ig VEE t + OSU'S CTw CIEE or re std Meteo eee orrerarnws views sores tn ———— adn tintin XO OS ATES

niin rte toch 1 menoepncrende\ OT HONE errs fetes oF ee papertninettintiyess (p +9re Fr wwendwe indethdiptndetadadada ainda’ . adadedads danendaainded et i ied ee ee iid openers
were Hee eeter se rere pytrtates: - ee ee Ta ere Ehsewio wre + * . or rr ewe OOO CU OTs ERT PETHT PTT eS rye erro et Tes Terr ree

cre peetetent any ere renre ne re _ . , or ee ee vewe OOF OO Re eww © ee Bee ee Sadbsameutainebeaen . — OPT ST

7 4 verre ges AAA wore om 6 RE Oe FOO FPPC SPSS OOTP Pr rw “- wre - oe _ mee See
; a SS STCureus ore ee

rowers TO Te Pee Te TeNTT EreT eee * popes wen -

ere eheryre arr ©
ree ee eee © e oF Ow Deere PHOT TET Ceeeree res ve

aed °
Tere eh ewe ee OFF OORT

Vere eer er ee me

eres Hote oe + ore "
oe eee ee Ow

ven
Ce ad 7 : . ;
‘“ teoe ree ee oe Oe © oe Hee ewe” verre hee wm ov vee
Opie otatete ee he's © : see = ~- owe r Cheers OF Fe OO Oe Oe Ore nee . eee-o ~—
’ . an at wena et a ’ sans em TU © OTe OK ome eee + (tor og eer peveeunuteares ° owwverewre
- ‘ °* i ° oe ” . . - = ~ ~
ere ’ vovw “er ao ed ie a Aon meen enstne ~ sranan ng oi nee oe eee

. a roe wwe
. mans

ere rrr

ee en

revere rere tee
og wre wr rows e

errr rrr ere

. re w etete et eer OF ©

re ewe ow
" on a, tind > Aldean lati Miia ciadh et ete hc teatiatct Deiat eect Dad ace alietindied .
TO PONS OV RT re enters DO oe wer ee geeury Pacieincie nents

Py Twos
oon oe we

err etrw rere

vier es etwets
poerretel eleva ertte wowre® &

es ee ted

tens 6 vrs
“oe i “"Trree e s*rrewe °
r . _ * .
one ~- > sree
* , : popes eens eet = 9a arm mc ded A hahah he ra eer PERT RES UTT? Oo
otyvrere we ¢ “ ° litle Ane hia “ vee ee ve oe ere me s ery” Pence Pees - . a eunerer - ——— rower
spokane wiereny arene ee eww ee forrrerre rr rPerere vee Vo re HUGS HH lee ewre. os eVEUTUTE CHET erepqumvercee eyorere eee HONE CUT ETOSTT OCCU Se
“~ . ory ‘ ’ ° Ae Bes vewe " . : Te OF ere wor wer” a > .
veer ees Ferere weee FSR “os ve aeerrenyrs sapetingioe exgmmeeteubresnan Se eeeeie en eens Owe FOO TE OWES CT COO EE OR EUE DCLG TENS
ns myer ewes ww orerete tt” owe were ,ourwwrsre roe paretetete eters Oe OE OE PFE FFF ere oe eee eee reer fee eres er! Gree HTT"
a — 7 ss . . 4 ida tia ees , erxe —_ 2 ~
‘oewr fF ree ree . we rwetes veer ewe ‘ “se “wes —_ ee « rrr fe wee o~ on
’ » ~v ' : . ~ oe ; a.
" rere 2 onw' pre” Sarooatieaed tena mane ores 9 OWT Ow? WEF . i mpen rerery rer? FOUTS TT PPR W re es coger es “Tere we EH HONT wee er™ Ooo
adng*erenengte rower erenerprenete eters pare . eee 1 Oe € Pere Te MO Rs Hit Rew E - a ro ft. +e OWNS FST ETY FORE UNS FEV EUY TEV UE RTO UWF cstandain tin nian
4 wrehen ov ' as oe wor ety des u | = . “ ~ eth .* re eee +e e ° coewre “*
an wanna ye on ~~ ve rw vir we” _—< ’ re s we ’ ower PT HT OT Pewiwry Pewee hl Pwegee wens “- were aes eet ene: qoou"
oe ew ' a) ‘ovr reece : ‘7 vv r tere Pe eT Fee ewer re re wr? . - ’ . . = ore
ores" . , ° - an eee. eee ee ee ee 2 7 ee ae s
(ve hae nena 4 ertsagm ree ; SAAR Adnan Sewerage ware er rery- # $0 91ee to omecgu em epee wererree ee PUR SS weae pew se Or eT TOTS OTT TOCTUT MT OS
: tuba . > ‘* ° . . . oor oe e aon - .
een is “~ : > Cote paipetiaiiibeeeeinee CRETE TN UOT RET FFT ETE Ge PETE ENE eer cer ese Tere Cee eTTT FE hEr TreT rT ess
we r . ve > ; ‘ - , . °
eee | 2 anon eating eal repens ”” lh cl led ak te et tnt ett ta tee Fee wer OCrs SWC UT OES COEUR UT ET CTTS
oe -ereeww ren ewer oes ate — Terr rr err TUN | Peres ee ev erwewrw ree . aad
ee! 9 'e% . wre : nv che Pe wow vee ee eee oe ea _-
- : . ' verrwys +) wee oe 7". . . evewveetrewe . Z -
on tae som oon qn ter pe pln errs et tn ephedrine intents a wAeves bn dhepitaintindh-tinsinsetrs nang ee eT POU COS FS wet FT OT ee ee oweewst oe . ove erere
a veers ee NTO etietiyt eT ee ee rors ee 9 teens Seinen ee WHET 2'O 8 CbCE COUITV CTY 5) Sveurerr rere ror
“A . . . a owes r v wer . “ ~~~
sere errr oy PERNT See rey eerer ees errr es ore footie A ae + Soles me tot — . as TOU Ce Et OF OD VOR SW wwE TF ewes OOO ee goth oe Stee wee ee wee erro wwe”
ainda poo ; omrawwrres SST asl enianre gree . ETOP Ore me OWT ETT Hee PEW EEF OW ewe eerreerer rrr rrr”
- verve ter owe ° : Se ee ee ee ee eer err esrer res eerere rere
ne owes ve bre ‘ ve rere eee te veer — ee eetererte : rarer
oer te ewe OW EF ited pew we ort eww eee ° .
ee ew OTe Oe OF ' , vve eT ee ee 0 ore yeas foo eee er? Pwew”~ ow . .
orn eee eee gerwerens eo vee P etnw w we wr ee . : A ae reer eRVece were ee ee ee ed As
yore weereeniyyte Wir vee tee e ere'v" oe vee ewnret ¢ whew eee : m ere CREE ETS Cow ere e ree vee re owen « - aa te 2 A mmm tt hac ° -oo-
ve oe ow : Dae > oweeer : ° ‘ : . 4 - “ -
lee ~~ ounen “ pate ae . asietuaranedl “er pe ta i cpt ws : owerrwrre ecene a. elif atincintinlind a —ee
: ’ . be ’ ee ee ~ore . ave Gptieds . verre wees ~e ~—oreorwe
wv : ee ‘owe ~ a ’ : ° - eee oor or ,
Ly ub ato fate et sh etta ade wo vn neote PewErenweesenreT Tres ts | eh rr eweER EN 1 ee eee eee e™ i ee ovovtowyos Pa ~ = ae mre a | a te thane se Fewwe
“y , ‘ vow vie rev poh 9 ae ie Owe errete ® ahew ee whew ee ¢ retro’ OF OR HHT Poe ee oe ee wy evr ep erry stews. os meee _
’ sews weer re we ere ’ ated ein into Star nsnee PP" POV erV RT ERs wrt rr ewe ese . ree ewes FECT TEE Heer eee ee .
~—'e" ees . e ef . ; - slp ietindintindind eee wes < er wr
vverewe'rr PVETVG tee erew rrr re were’ vrewrrrwrerr’ a ee ee ee . ceduerrwre ou — ee anv _—a vor vere
wrtew' ove eye rreer sy reese ryerrwrer rrr’ * nd didi ’ rer tee were oes : be te ha wr rer
? eres ' ’ - “aeveewr . sever te» . ‘ ry 7 . . pears oS
. oie ane reyes eoyerreen yo See ° shbees a0 . ** . —_— ~ 7 ~- wre AA Ae ee ee ee
~~ a tne OT ea diel ; +e wore Kee es _ seee wower pureenwesn -
“er vewvreerwrve® ee - eens e~ or : retvremwvrwr ? a ’ . ¥ i ° e
rl a . wastes ere eee ** ewer teewe * ve +e Pw ee WPT Ew ease tt ma
~ ° ? rw oe - voor? errors oe wee wre re reer ’ r a " . oer ererte : .
are » alin “* ¢ ’ Z ’ rrhrer were ° el el rT tree . “ os eee a <te
a . oO rere ager" 4 . ve : - . : ° ae oe mand ure ae > . ‘
a A eee oe eprmned eteve” ve Erourneneeye nos om vier sl : orev envs CONF Tee wer rrent
. sige —t fa Oe owe oye oe sntabded POSS > ® vretee ee ee ee pewrrere [oe c oer o% ood ¥ evveee - “ owe ovo
yerrwrrrT? mm wing ’ if eve’ eet ° . . vw ~oer corre ree bey eRe © PWN Hee ee Pew A ae sills ° . - r ——
° ll - ee ’ A > r ‘ = ‘oy . le ° ¥ ial a
oor ’ ° _* ° snc vrewe ee i eee ee — rues
nentuatere® a atta lit stele thd ¥ renews Cre rv eew « .
. te . i e A * fl PO Hee Howe . pore 1 he oe eo o~ ee ~~" i
dnl shirt ln an . : sled ' ' . vv rer rlrrer ree 8
ewes - . es vee srae : } dneintiol Air tetn atin gos eee
rege eee eee eee SENT OEP r 7 : . , mh se vwe
SA AGS omeret al ji fet - “ owe ees "ve
7 . . * A : Re . ’ . . : mrere v -«
we vr “a be on « ' Crvnerrr errors ans
mw oe ’ ° . far fn tt e omer y owe re ee oo *
’ ove X - aol . ;é " “wr
‘eo

oo“ -
prrrervrrret*
wonwerrrr’

_verrer

rer? POOR Pew ree ow
, et errr ee wee °F Oe ew ew wee
ee ee ee ee ee ee ee en **
Firetree oer ee LALA ee ee ee
1 eee rere
:

peer er ener ‘ te | ‘ ‘ > . . eutewe tereea eer

weerrrreers’ tees
= Sore Le a Ae
Pe A A
beer POW Re FF eee ere
eee et ee Ae : te

errs FOR Per ee , ed
COT PP ORT Ober eT wee ee

err
fret eter

+" eevee

Vee Or eee PTET TO PROT were i
DROS RAS eee eee ee. ee ee ee
eter eo a eel “vereverve
ee eee eee ee FOOSE HH eH Oe Owe Oe ee &e
TP Pee ee” ce ee Fe ee OP? Oe! OO ee Fe OTe Owe ws © ee
PO" CHF PTO CR es HT PET eT TTT eT OP res seewerer:
Pee OO COMERS ete es +t eee we ee ee coostous eee rerrre?
(vee a ne enn es aa
Se dee dete a eh ee te LL nd hh heh ett
OE TS SPS OOO Re Orr eT ews Cree

~ adbdbcaghdin ttn: sutintentndndedne, bs Crews

rere

somrwe — eters tere erew ne re

C8 FO Wee ees ee oe oe Se ONY FRO ORR ee
TeV Pre eee ree rowrre tr ore
or" ; CeO ORE T Eee Oe Oe ee Her
~ » e 4 Pe es a lh eda ele
feeeene Gane ; —— wm han r oA .

. owe? severe
te anger eese ge eee

’
7°

En ee

reer ee

weetwrr —
onawrene oer
—_ "> a)

-eeeee
>

a =<
‘omens. som = ‘one Wy Be
PAPE SBEL RET ON UTERO Y any
Ay Tal yAn Ty ers tod y 4 Sas oes
*-- we wy : = Pin paviiniys ERC paenadn FW Nw af
{ | iN » preigit aise
Vite I Mid '

Negril
ri
\ I MAA.” “

vege” wy apea on
ay! Orth fnew
He monte

Leeman aeons
| ah Anag. TORE anee | |)
w lili JUTE CONAN AEA ana? Bideldel PL Ltr Keg pit vy ivy | Val

Hii ee he Hlhnijanminngtig wean Wai
E iy

toe

wip idl abbadds

~ 0a

ial al al Oy

PAAA- wht AAAnm
i 4
wty~e fs

Wakicsns Mee 1 | wl aN
Veen “Nay fs Pitvewars, x
Pr eenn” % VeMAAS. CAS Aiko, Pree wy | : &
ee ‘Aw / : 4
. ; f d af Shee penton vy ' Ny NAY ° Piet 4

é
yin ue we
“ st Naan

ee Noten
SANNA MAAS anne

Dia ees wet” th, Vas watacanen 1] THT re Titel dd
Yd ee: i; ee Set,
bn Devnet al HU

t hed | Ly

: nti Xi seller ta i

N es y ¥ veces 1c ie | Ad |
hve a peer || hee

(i

ores

Misco mean ay
srl poate ss, Neat np 9 Ong
Ad | Ves ° mt are

g = : PAL
ad Daet 7 ion H eee

dea MN v rid LT ELPEE PET aoe? oe | Whi Hd

<A
, is S : " | 4 wt tAdbAAn,, j ; +5 Pee \ ~
LY exe, Af * 4, am a oe | | A | Ave 4 sai lavmagnlet Femmes nd \ "\. Ww :

bal _ “3 . eye ae ;

SPE Oe sifwy fnr- . Mt Ayes * AA, a xy
lid 4. Ty Ar AI oe 5 ee ye alot ol il el AA ay I \w. ‘ verre. |
Jay a“ ser"... ‘ ] se “4

Tt fyi 1
vs yi “verte Ane A nn ny “\A Hrs : if

| | m<s ae wat
i Wes gry at? ° ae b * dunt ‘ “~ ,
», “™ p a? ‘ ss ~§ bn ! s >

ON 9° “vw 1S Wx ue be § wee’ oF hs - Tuer
=f rawant Bee eeeeeeran Ws peri
si & ya . |

’ » ; "te . «94 a

tn yn EEL
jm, Ses / ‘ayaeee ee uses f\ eH ‘ Bice’t “s Isa peee’

al . 5
a oe Me ~ AAA LS a , | Beg) Yay
Ti adie Wa we Ta tiaet TTT TH | a ae Art, ee
os ) ey, . ie retin, { i is bat yh abot Rh 20 ; A “ Te
wie | Mrtyne anti de! UM tvrne. wu XY “be eyes- We i 1 we te MARS Pi

a \v Na)
NN Ag ty) | h “5b Vitrraam: Ape LLL tatttenenagtts git ay “Wiivig!ing +e
| vy via’ we bik TTT TTT ae ancah V4 7m Ne, ed RES Pr'ny dail

ty Pe i OGD | |, Phiten
Ly Ops A i Kye’ . 7 ed ld ats a | hie wa, \4
Morden coe oe iyntttnt Meagan WL ald J i eee
| Tiel TR Cd LL Lett NO ee Nee Lae Pow at cogig ht! Ting wi Nek | wen
ftw. fo, Ad) enna itl SaaS eT ; nee A Ae wy ’ iy
Perel ct altel Male 1 MOO | ir SAG Crna ay Eel andl
wow Vella seenre ON, Np Ay ." Mi. r | ' nal ' Ky ei vy Den Lint ie |
et bi! ALL ihe ts a | Tltlen tt 't ny vest Uo veae ee Teas a.
Le DOMES CET Mia ean li aid lnleiial RoR dei aL ALAS la NMI aD I) BNR Petal
elk ay hae ayant pr ee | wer whe ) ded belied |

ins, bang HARP . {PUA LOL £0 Miars iy wry
wee ry sonst wv mf a Tl Ut he he ‘ vapengetenn® Ml Pie? reat |

vr, r<
me re! Ary A veers wap A Nes ms
; anne: PTT Lele DIT Sah EDR Lh wall | TERT,

of SUT 1 ahd, snahe"in Nl
adie aN ree ey pert ONAN 2 Pitan mA QD. bd

3G

AMERICAN
JOURNAL OF SCLENCE.

Epitrorn: EDWARD S. DANA.

ASSOCIATE EDITORS
Proressors GEO. L. GOODALE, JOHN TROWBRIDGE,
H. P. BOWDITCH anv W. G. FARLOW, or Camprines,

-Prorressors A. E. VERRILL, HENRY 8S. WILLIAMS, anp
L. V. PIRSSON, or New Haven,

Proressor GEORGE F. BARKER, or Pamapetpnia,
Proressor H. A. ROWLAND, or Batrtimore,
Mr. J. S. DILLER, or Wasuineron.

FOURTH SERIES.

VOL. VII—[WHOLE NUMBER, CLVII.]

WITH ELEVEN PLATES.

we A, : ; :
NEW HAVEN, CONNECTICUT, SSwepal Museu

1899.

THE TUTTLE, MOREHOUSE & TAYLOR PRESS,

j

NEW HAVEN, CONN,

‘
¥S

ee
“

i ee en

T= - ce"

CONTENTS TO VOLUME VII.

Number 37.

Page
Art. L—Thermodynamic Relations of Hydrated Glass; by
BEET otters! Soy 5 SS Se sie eee 3 kl 1
I].—Platinum and Iridium in Meteoric Iron; by J. M.
EE aS a OS <2 ates eee 4
Ifl.—Studies in the Cyperacee; by T. Hoim ___---.-._--- 5
IV.—Regnault’s Calorie and our Knowledge of the Specific
Volumes of Steam; by G. P. STaRKWEATHER ------_-- 13
V.—Estimation of Boric Acid; by F. A. Goocu and L. C.
ee oe. Oe oe bs eee h tee BA
VI.—Descriptions of imperfectly known and new Actinians,
with critical notes on other species, II; by A. E. Ver-
nS... 5 nee ee re OE Se 41
VIl.—Mineralogical Notes; by W. F. Hittepranp-..----- 51
VIII.—W hat is the Loess? by F. W. SarprEson-_--._----- 58
1X.—Absorption of Gases in a High Vacuum; by C. C.
ERR fc = I ae pine 61

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Extraction of Nickel by the Mond Process, ROBERTS-
AUSTEN: A®therion, CROOKEs, 64.—Preparation of Graphitic Acid, STAUDEN-
MAIER: Compounds of Lithium and Calcium with Ammonium, MOISsAN, 65.—
Spectra of Iodine, KoNEN: Manual of Chemical Analysis, Qualitative and
Quantitative, G. S. Newt: Stratified Brush Discharges in Atmospheric Arr,
TOEPLER, 67.—Spectrum of Lightning: Dispersion in the Electrical Spectrum,
K. Marx: Traité Elémentaire de Méchanique Chimique fondée sur la Thermo-
dynamique, P. DusEeM, 68.—Prismatie and Diffraction Spectra; Memoirs of
Joseph von Fraunhofer, 69.

Geology and Mineralogy—Maryland Geological Survey, 69.—Lower Cretaceous
Grypheas of the Texas Region, R. T. Httn and T. W. VauGuHan: Bibliographic
Index of North American Carboniferous Invertebrates, S. WELLER, 70.—Con-
tributions to the Tertiary Fauna of Fiorida, W. H. Datu: Contributions to
Canadian Paleontology, J. F. Wuiteaves: Geological Survey of Canada: Re-
port on the Doobaunt, Kazan and Ferguson Rivers and the Northwest Coast of
Hudson Bay, J. B. TyrreLu, 71.—Rivers of North America, a reading lesson
for students of geography and geology, I. C. Russeiu: Earth Sculpture or the
origin of land forms, J. Gerkiz, 72.—Elemente der Gesteinslehre, H. ROsSEN-
BUSCH, 73.—Educational Series of Rock Specimens collected and distributed by
the U.S. Geological Survey, J. S Dimter: Mechanical Composition of Wind
Deposits, J. A. UppEN, 74.—Brief notices of some recently described Minerals,
75.

Botany and Zoology—Poisonous effect exerted on living plants by Phenols, R. H.
True.—Peculiar mode of formation of pollen-grains 1n Magnolias, GUIGNARD:
Elements de Botanique, P. Van TieGHem, 77.—Sketch of the Evolution of our
Native Fruits, L. H. Baitey: Bush-Fruits, F. W. Carp: Metamorphosis of
Asterias pallida with special reference to the fate of the Body Cavities, S. Goro,
78.—Zoological Results based on material from New Britain, New Guinea,
Loyalty Islands, and elsewhere, collected during 1895-97, A. WiLLEY: Account
of the Crustacea of Norway, with figures of all the species, G. O. Sars: Mol-
lusea of the Chicago Area: The Pelecypoda, F. C. Baker: Catalogus Mamma-
liam tam viventium quam fossilium, E.-L, Trovessartr: Fishes of North and
Middle America, D. S. Jorpan and B. W. Evermann, 79.

Miscellaneous Scientific Intelligence—Institute of France: Annual Report of the
Board of Regents of the Smithsonian Institution: Ursprung der Afrikanischen
Kultur, L. Fropexius: Organic Evolution Considered, Farruurst, 80.

lV CONTENTS.

Number 38.

P
Art. X.—Contribution to the Study of Contact Metamor- e
pinsms ‘by J. Ma°CLEMEN TS! 935.2 k ote ee eee 81
XI.—Origin of Mammals; by H. F. Osporn ._..-_..--_-- 92
XII.—Chemical Composition of Tourmaline; by 8S. L. PEn-
pintp and’: \W. oOore 2.455 a a 97

XIU.—Littoral Mollusks from Cape Fairweather, Patagonia ;
by HB. A. Pitspry, “(With Plate I.) .- 222523) 126

XIV.—Thermodynamic Relations for Steam; by G. P.
STARK WEATHER) Gah. 1) eee eos BL 95 5) sr 129

XV.—Descriptions of imperfectly known and new Actinians,
with critical notes on other species, Ill; by A. E. Ver-
BTL 2) oc easements SS SED Wile ot aa 143

XVI.—Volumetric Method for the Estimation of Boric Acid;
by Ly C. Jonns 2202 o>)... Br

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Molecular Masses of Gases, D, BERTHELOT: Preparation
of Metals and Alloys by means of Aluminum, GOLDScHM1DT, 154.—Electrolytic
Preparation cf Beryllium, LEBEAU, 155.—Colloidal Silver, LOTTERMOSER and VON
MEYER: Silver peroxide and peroxynitrate, MULDER, 156 —Neodymium, Bovu-
DOUARD: Lecons de Chimie Physique professées 4 l'Université de Berlin, J. H.
Van’r Horr: Outlines of Industrial Chemistry, F. H. Tnorp, 157.—Charge of
Electricity carried by the ions produced by Rontgen Rays, J. J. Tomson: Use
of the Coherer in Measuring Electric Waves, O. BEHRENDSEN, 158.—Color

Blindness and the Réntgen Rays, I. Dorn: Regenerating Vacuum Tubes, W-.

ROLLINS, 159.

Geology and Natural History—Age of the Earth, Lorp KeEtvin, 160.—Recent
Publications of the U. S. Geological Survey, 166.—Catalogue of the Cretaceous
and Tertiary plants of North America, F. H. Knowiton: Iowa Geological
Survey, Annual Report, 1897, 8. CaLvIn: Preliminary Report on a Part of the
Gold Deposits of Georgia, W. S. Yeatzes, S. W. McoCatiis and F. P. Kine, 168.
—Report of the Governor of Arizona to the Secretary of the Interior: Die
natiirlichen Pflanzenfamilien, ENGLER: Fishes of North and Middle America,
D. S. JoRDAN and B. W. EVERMANN, 169.

Miscellaneous Scientific Intelligence—Ostwald’s Klassiker der Exakten Wissen-
schaften: University of Tennessee Record of Scientific Engineering, 170,

—

CONTENTS. Vv

Number 389.

Page

Art, XVII.—Studies in the Cyperaceez; by T. Horm._--.- 171
XVIII.—Granitic Breccias of Grizzly Peak, Col.; by G. H.

0 OY se Sse es ae Se I See tee

XIX.—Constitution of the Ammonium-Magnesium Phos-
phate of Analysis; by F. A. Goocu and M. Austin .- 187

XX.—Crystal Symmetry of the Minerals of the Mica Group;
mleerite ny Anon sy: teem ee eh 2 199

X XJI.—Descriptions of imperfectly known and new Actinians,
with critical notes on other species, IV; by A. E. Vzr-
SAE Ss A ae ae ES geet A Gy Ape oe geen Pa ee 205

XXII.—Study of some American Fossil Cycads. Part I.
The Male Flower of Cycadeoidea; by G. R. Wisanp.

eiveith de lates: DE-DV..): ._'. | sete yt SHS 219
XXI1.—Footprints of Jurassic Dinosaurs ; by O. C. Marsu.

Meyrin inlate NN.) 222.2 Se Comme ee a a A 229
XXIV.—New Kansas Meteorite; by H. L. Warp.---._- 43238

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Density and Composition of Liquid Air, LADENBURG
and KRUGEL, 234.—Semi-Permeable Membranes, MisEers, 235.—Solution of
Platinum and Gold in Electrolytes, MARGUELES: Aqueous Solutions of Metallic
Gold, Zsia¢mMonDY: Explosion of Mixtures of Methane and Air by the Electric
Current, Cournot and MEuNIER, 236.—Crystallized Compound of Cuprous
Chloride and Acetylene, CHAVASTELON: Matter, Energy. Force and Work, 8.
W. Hotman, 237.—Uranium Radiation, and the Electrical Conduction produced
by it, RUTHERFORD, 23&.—Silver Voltameter and its use in determinations of -
normal elements, K. KAHLE: Absorption of light by a body placed in the mag-
netic field, RHIGI, 239.

Geology and Mineralogy—Recent Earth Movement in the Great Lakes Region,
G. K. GILBERT, 239.—Petrology of Rockall Island, J. W. Jupp, 241.—Recher-
ches géologiques et pétrographiques sur le massif du Mont Blane, L. DupARrC
and L. Mrazec: Native Silver in North Carolina, G. F. Kunz, 242.—Occurrence
of polycrase in Canada, G. C. HorrmMann: Kaolins and Fire Clays of Europe,
and the Clay-working Industry of the United States in 1897, H. Rigs, 243.—
Origin of the Gases evolved on heating Mineral Substances, Meteorites, etc., M.
W. TRAVERS, 244.

Botany and Zoology—Symbole Antillanz seu Fundamenta Floree Indize Occiden-
talis, I. URBAN: Plant life, considered with special reference to form and
function, C. R. BARNES, 244.—Rhodora: Journal of the New England Botani-
cal Club, 245 —Catalogue of the Lepidoptera Phalenz in the British Museum,
Vol. I, Syntomidee, G. F. Hampson, 246.

Miscellaneous Scientific Intelligence—Report of 8. P. Langley, Secretary of the
Smithsonian Institution, for the year ending June 30th, 1898, 246.—Physical
Geography, W. M. Davis: Annual Report of the Director of the Field Colum-
bian Museum, 248.

Obituary—Louis W. PEcK, 248.

al CONTENTS.

Number 40.
Page

Art. XXV.—Glacial Lakes Newberry, Warren and Dana,
in Central New York; by H. L. Farrenizp. (With
Plate Vil) o cee eos. 2. ee
XXVI.—Rapid Method for the Determination of the
Amount of Soluble Mineral Matter in a Soil; by T. H.
MRmaAns i 2. flu ea ll
XXVII.—New Type of Telescope Objective especially
adapted for Spectroscopic Use; by C. 8S. Hasrines._-. 267
XX VIII.—Phenoerysts of Intr usive Igneous Rocks; by L.

Vi PinssOn Poe ee ee 271
XXIX.—Occurrence, Origin and Chemical Composition of
Chromite; by J.-H. Pratr: 2002s) 281°

XXX.—Influence of Hydrochloric Acid in Titrations by
Sodium Thiosulphate, with special reference to the Esti-

mation of Selenious Acid; by J. T. Norron, Jr.__.--- 287
XX XI.—Rock-forming Biotites and Amphiboles; by H. W.

PURNERie! a8 5 Ger... > ee a
XX XII.—One little known and one hitherto unknown species

of Saurocephaluss* by O. Pasay 22. 322502 eee 299
XXXIII.—Some American Fossil Cycads. Part II; by G.

Re Wrevann,- (With PlatesVIL) -2°. 7.22 2a 305

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Luminous and Chemical Energy, BERTHELOT, 309.—
Homogeneity of Helium, Ramsay and TRAVERS: Properties of Ozone, LADEN-
BURG, 310.—Preparation of Alkali Nitrites, Divers: Composition of American
Petrolenm, YounG: Electrical waves along wires, A. SOMMERFELD, 311.—
Polarization and hysteresis in dielectrics, W. SCHAUFELBERGER: New method
of showing electric waves, A. NEUGSCHWENDER: Separation of long waves of
heat by means of quartz prisms, H. RuBENS and H. AscuKinass: Theory of
conduction of electricity through gases by charged ions, J. J. THomson: Hall
phenomenon and the theory of Lorentz, PoincaR#, 312.—Relation of the Sur-
face Tension and Specific Gravity of certain Aqueous Solutions to their State of
Jonization, KE. H. ARCHBALD: Cause of the Absence of Coloration in certain
Limpid Natural Waters, W. SprING, 313.—Text-Book of General Physics for
the use of Colleges and Scientific Schools, C. 8S. Hastines and F. EH. Beaca, 314,

Geology and Mineralogy—New facts regarding Devonian Fishes, C. R. Hast-
MAN, 314.—Geological Sketch of San Clemente Island, W. 8. T. Smita: Geol-
ogy of the Edwards plateau and Rio Grande plain, ete., R. T. Hitt and T. W.
VAUGHAN, 315.—South Dakota Geological Survey, Bulletin No. 2, J. BE. Topp:
Crater Lake, Oregon, J. 8. DILLER: International Geological Congress: Inter-
national Geographical Congress, 316.—Funafuti, W. J. SoLLAs: Petrograph-
ical and Geological Investigations of certain Transvaal Norites, Gabbros, and
Pyroxenites, J. A. L. HENDERSON, 317.—Occtrrence of Corundum, T. H. HOL-
LAND and W. G. Miutmr, 318.—Minerals in Rock Sections, L. Mol. LuQuer:
New Mineral Names, 319.

Miscellaneous Scientific Intelligence—Benjamin Apthorp Gould Fund, 320.—An-
nual Report of the Board of Regents of the Smithsonian Institution for the
year ending June 30, 1896: Second Washington Catalogue of Stars, 321.—
Zoological Results based on material from New Britain, New Guinea, Loyalty
Island and elsewhere, collected during the years 1895-97, A. W1LLEY: Select
Bibliography of Chemistry, 1492-1897, H. C. Bouron: Biological Laboratory,
322.

CONTENTS. vil

Number 41.

Art. XXXIV.—Some Experiments with Endothermic
Pere Oy NYG. MerER | 24a ee. 823

XXXV.—Hypothesis to explain the partial non-explosive
Combination of Explosive Gases and Gaseous Mixtures ;

Page

PG ONMINP HIS ce ee i ok OF
XXX VIJI.—Occurrence of Paleotrochis in Volcanic Rocks in
: medica oy ES. VV ILIAMS oe jhe ore es: 335
XXX VII.—Origin of Paleotrochis; by J. 8. Diturr _-.--- 337

XXXVIII.—Association of Argillaceous Rocks with Quartz
Veins in the Region of Diamantina, Brazil; by O. A.
iprree e k . . a ee e ea tne 343

XX XIX.—Goldschmidtite, a New Mineral; by W. H. Hopss 357

XL.—Hydromica from New Jersey; by F. W. Crarxe and
ela SIU ABTON $222) .iu. 1d 2 Sa 2 eee) SGD

XLI.—Powellite Crystals from Michigan; by C. PatacnE__ 367

XLI.—V olatilization of the Iron Chlorides in Analysis, and
the Separation of the Oxides of Iron and Aluminum ; by
Hee Goon andl. §. Havanese 370

XLUI.—Descripiions of imperfectly known and new Actini-
ans, with critical notes on other species, V; by A. E.
> ETE TR TTT pg gaps es AO nM 2: eg acl 375

XLIV.—Preliminary Note as to the Cause of Root-Pressure ;
Pye. -G. LEAVITT. _! 381

XLV.—Study of Some American Fossil Cycads. Part III;
by G. R. Wietanp. (With Plates VITI-X.)..-.----- 383

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Low Temperature Experiments, HEMPEL, 392.—Reduc-
tions in Presence of Palladium, ZeLINSKy: Properties of Metallic Calcium,
MOIssAN, 393.—New Bases from Strychnine, TareL: History of Physics in its
Elementary Branches including the Evolution of Physical Laboratories, F.
CAJoRI, 394.—Electrolytic Interrupter, A A.C. Swinton: Measurement of slow
electrical oscillations, W. Kovic, 395.—Electric waves on wires, W. D. COOL-
INGE: Absorption of the X-rays by air: Optischen Instrumente der Firme R.
Fuess, C. LEIss, 396.

Geology and Natural Histovry—Note on a Bridger Eocene Carnivore, O. C. Marsh:
Age of the Franklin White Limestone of Sussex County, N. J., J. E. WOLFF
and A. H. Brooks, 397.—West Virginia Geological and Economic Survey:
Alabama Geological Survey, W. B. Puruties: Development of Lytoceras and
Phylloceras, J. P. Smita: Pre-Cambrian Igneous Rocks of the Fox River Val-
ley, Wisc., 8S. WEIDMAN, 398.—West Virginia Geological Survey, I. C. WHITE:
New and interesting olivine-melilite-leucite rock, V. SABATINI, 399.—Deuxiéme
Mémoire sur les Algues Marines du Groenland, L. K. RoSENVINGE: Mono-
graphie des Caulerpes, A. W. Bossr, 400.—Dispositione Delesseriearum, J. G.
AGARDH, 401.

Miscellaneous Scientific Intelligence—National Academy of Sciences, 401.—Annual
Report of the Board of Regents of the Smithsonian Institution, to July, 1897:
Harper’s Scientific Series, 402.

Obituary—Dr. GUSTAV WIEDEMANN, 402.

Vili | 7 CONTENTS.

Number 42.
Othniel Charles Marsh; by C. EH. Buecumr _.-. 72222 eee 408

Arr. XLVI.—Camden Chert of Tennessee and its Lower
Oriskany Fauna; by J. M. Sarrorp and C. ScuucnErtr 429

XLVII.—Recent discovery of rocks of the age of the Trenton
formation at Akpatok Island, Ungava Bay, Ungava; by
Je EF. WHitmaves ...-. 222000 bo. 22 13 Ses

XLVIII.—Studies in the Cyperacez, X; by T. Hotm_----- 435

XLIX.—Roscoelite; by W. F. Hittespranp and H. W.
TuRNER, with a note on its chemical constitution by F.
We (CLARKE yh wee ee op ee eee ee 451

L.—Gravitation in Gaseous Nebule; by F. E. Nipuer ._.. 459

LI.—Titration of Oxalic Acid by Potassium Permanganate
in presence of Hydrochloric Acid; by F. A. Goocu and
C.. A... Parers oo... 2... 92s 2S 461

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Influence of Electric Oscillations on Vapors. KAUFFMAN:
Occlusion of Oxygen and Hydrogen by Platinum black, Monn, RAmsay and
SHIELDS, 468.—Certain Derivatives of Acetylene, ERDMAN and KOTHNER, 469.

.—KEutropic Series of the Calcium Group, EPPLER: Distribution of Vanadium,
HASSELBERG, 470.—Determination of time of vibration of very high notes, F.
MELDE: Heat conductivity of different kinds of glass, A. WINKELMANN: Ab-
sorption of Electric Waves by different substances, E. BRANLY and G. LEBON:
Radiations of Uranium, H. BEcQuEREL,, 471.—-Source of Uranium Radiations,
W. Crookes: Phosphorescence at Low Temperatures, A. and L. LUMIERE:
Recueil Données numériques publié par la Société Francaise de Physique, H.
DuFET, 472.

Miscellaneous Scientific Intelligence—Icthyologia Ohiensis by C. 8. Rafinesque, R.
EK. Catt: Elementary Physiology, B. Moors, 473.—Experimental Morphology,
Part II, C. B. Davenport: Recherche, Captage et Aménagement des Sources
Thermo-Minérales, L. Dr Launay, 474.

Obituary— Dr. FRANZ RITTER VON Hauer, 474.

INDEX TO VOLUME VII, 475.

Chas—D. W alcott,
__U. S. Geol.

U. serve J /be ve

fe: AMERIOAN — |
OURNAL OF SCIENCE

‘Eprror: EDWARD S. DANA.

ASSOCIATE EDITORS

4

ey. ©. ‘MARSH, AS eas I VERRILL AND H., 3,
WILLIAMS, OF ae Haven,

ee H. A. ROWLAND, OF 7S Pao
. Mr. J. S. DILLER, or Wasurneron.

FOURTH SERIES.

VOL. VII-[WHOLE NUMBER, CLVII.]

No. 37.—JANUARY, 1899.

NEW HAVEN, CONNECTICUT.
Ei eeg So ee

Xe * uf
’ at » .

ae mOnEROUSE & TAYLOR, PRINTERS, 125 TEMPLE STREET. .

ablished monthly. Six dollars per year (postage prepaid), $6.40. to
gn subscribers of countries in the Postal Union. Remittances should

: os de | either Aes baat’ aa PEGE ee: letters, or bank see

RETAILING CF THE FAMOUS

TRAUTWINE COLLECTION.

We have just purchased entire, one of the first among great American —
collections, in point of species, sub- species and varieties represented i in excel-
lent type specimens. It shows few efforts at display, but is a gathering of
over 6,000 specimens, including hundreds of rare numbered species not
found in the average collection, ‘and nearly a dozen which are to be seen on

- the desiderata lists “of the most ‘complete American and European collections.

John C. Trautwine, Civil Engineer and Mineralogist, was born in Phila-
delphia in 1810. He began the collecting of minerals about 18385 and con-
tinued until his death in 1883. The collection was built up through
exchange, purchase and extensive foreign travel. It may justly be said to
cover the period of 1840-70, illustrating with wonderful completeness the
mineralogical discovery of the time—so well in fact, that the collection was
studied by nearly all of his mineralogical contemporaries in America. Much
of this discovery is found recorded to-day only in the nomenclature of the
science ; good examples are largely foreign to recent collections. Types and
localities are shown in great variety under all important and many obscure
species. Labels are models of scientific accuracy and neat penmanship,
being small and attached to specimens. There are many duplicates, other-
wise it would be useless to publish a list, as specimens will sell at sight.
The short list of more important things will be mailed on application.
Prices vary from about 50c. to $5.00 more or less. We know of no minerals
so-universally desired as the old historical ones, and they are here by the
thousand. Familiar minerals from the original localities from which they
take their name, are frequent. South American and European localities fully

represented,
ABOVE IS AN OPPORTUNITY

which is not offered collectors once in ten years.

A “PF RIP: ‘TORT Err

BROKEN HILL MINES, N.S. W.

Over 1,100 pounds of splendid crystallizations have just arrived as the

result of a five weeks’ trip of our special salaried collector. No more than

brief mention can be made here of the leading kinds. Of some a consider-
able supply was obtained and old prices are greatly lowered. They are far
superior to what has been on the market heretofore. Our illustrated supple-
ment to appear late in January will describe them fully—mailed free. Send
for it. The full page engraving of stellate twin Cerussite, coated with
Anglesite xls, is an artistic success and illustrates a grand crystallization.

STOLZITE, gem like xls on gangue. Rare!!

RASPITE (with Stolzite) gem like xls on gangue. New!! Rare!!!

‘ CERUSSITE. Masses of crystals i in six-rayed stellate twins; also “‘V”
shaped twins! !

ANGLESITE xls coating the foregoing. Some choice museum groups at
low prices, for this rare combination 1!)

EMBOLITE crystals and sponge-like masses, one-half former prices.

PYROMORPHITE, masses of bright, prown xls.

IOYDRITE, cryst. "Rare ! |

The Trautwine and Broken Hill minerals, together with other December
accessions, will be placed on sale J dene” Ath.

Dr. A. EK. HOOTE,

WARREN M. FOOTE, Manager.
DEALER: IN) MINERALS:

1317 Arch Street, Philadelphia, Pa., U. S. A.

Established 1876.

yl

THE

AMERICAN JOURNAL OF SCIENCE

[FOURTH SERIES. ]

Art. 1.— The Thermodynamic Relations of Hydrated Glass ;*
by C. Barus.

1. FURTHER treatment of the fusible water-glass described in
the preceding communicationt has brought out the following
points: During the first or opaque stage of the reaction of

hot water on glass (200°), volume-contraction Ge ) and increase

of compressibility (@) are both marked phenomena. During
the second stage the water-glass becomes more and more clear
and limpid. Volume-contraction falls off asymptotically to
zero. Compressibility, after passing through a pronounced
maximum which may be even larger than 500/10° per atmos-
phere, diminishes with great rapidity to the probable isothermal
value for pure water (about 100/10"). The relation of change
of compressibility to the corresponding volume-contraction
(per unit of volume, cold), i.e.d 8/(d v/v,), remains throughout
of nearly the same order, lying within an interval of 10/10° to
20/10°, and nearer the former limit so far as measurable.
The total apparent volume-contraction of the hot column of
water-glass may reach 30 per cent.

2. On cooling bubbles appear in the clear water-glass in great
number, showing it to contract on solidifying from the center
outward (centrifugally), like a Prince Rupert drop. The solid
water-glass is in appearance as hard and brittle as ordinary glass,

*Read before the Nat. Academy of Sciences, Nov. 16th, 1898.

+ This Journal (4), vi, p. 270, 1898; cf. this Journal (3), xli, p. 110, 1891. The

experiments of this paper ending with compressibilities as high as 200/105, relate
chiefly to the opaque stage.

AM. JouR. Sc1.—Fourts Sgrizs, Vou. VII, No. 37.—JANnuARY, 1899.
1

2 Barus—Thermodynamic Relations of Hydrated Glass.

from which it differs in refraction and density. Capillary
tubes at the end of the process have become glass rods. The
original thread of water is replaced by a solid core of water-
olass more than twice the original diameter.*

Nevertheless all capillary tubes with a solid core of this kind
break throughout their length in the lapse of time, however
carefully cooled. Cf. § 5.

3. An explanation ot these phenomena may be given in the
terms of the compressibility observed along the successive
isotherms for the different concentrations of glass solution at
210°. This at first shows a relatively small value, implying a
steep isotherm in a Clapeyron (pv) diagram. Thereafter com-
pressibility passes through a relatively enormous value (over 5
times the initial result), implying a nearly horizontal isotherm.
The reaction ends with even smaller compressibility than the
first observed, implying steeper isotherms than the initial curves.
Now although these isotherms are all at the same absolute tem-
perature, the water-glass is becoming continually more goncen-
trated and viscous. Hence the corresponding temperatures of
the isotherms in van der Waals’ sense are continually decreasing.
It follows that the reaction considered as a thermodynamic
process is amarch through the erdtecal region of certain phases
of the water-glass examined.

4. Inthe light given by J. W. Gibbs’s famous investiga-
tions, one may arrive at clear notions by adopting but two
phases of the water-glass, for comparison. Many phases may
coexist ; two are selected in the interest of brevity and called
phase 1 and 2, respectively. During the earlier stages of the
reaction (dilute water-glass) phase 1 is stable. At the end of
the reaction (concentrated water-glass, subsidence of volume
contraction) phase 2 is stable. Both cases correspond to steep.
isotherms. ‘Toward the middle region of the reaction (max-
ima of compressibility) phases 1 and 2 are mutually stable in
presence of each other. | Hence the horizontal isotherms
corresponding to the critical region are cut through in a march
from greater to smaller corresponding temperatures.

That phase 2 is really unstable during the first stage of
reaction is shown by the approximate constancy of the values
of compressibility, throughout intervals of pressure as high as
could be applied. For instance, at 185°,

Interval, p= 20-100 100-200 200-800 300-400 atm,
10°xB= 146 144 142 146

And fifteen minutes later
10° x6 = 188 176 201 189.

* Diameters of threads of water and of water-glass respectively in tube No. 4,
024 and ‘071°; in tube No. 5, 033 and °059™; in tube No. 6, (053 and 125°™;
etc.

Barus— Thermodynamic Relations of Hydrated Glass. 3

Again, during the intermediate stage, water-glass is non-resi-
lient: yielding remarkably to increase of pressure, it refuses to
expand when pressure is removed. The last stage is again
elastic but relatively incompressible.

The extreme compressibility of water-glass during the stated
intermediate stage deserves especial comment. Beginning
with igneous glass and water* with compressibilities of the
order of 3/10° and 100/10° respectively at 210°, values are
reached in water-glass which exceed 500/10°. One may even
compare this result with so volatile a body as ether at different
temperatures and between 100 and 200 atm., as follows:

Temperature = 29° 65° at O0° 185°
BX 10° = 156 207 305 741

Yet the water-glass solidifies in the cold to a hard colorless
body quite resembling igneous glass.

5. Made in quantities in a large digester, water-glass is
obtained as a nearly homogeneous compact body, adhering forci-
bly to the walls of the retort, from which it must be removed
with wedge and hammer. A small lump held above a candle
flame soon fuses with loss of water to a milk-white pumice.
Left without interference in the cold for several weeks or
months, it spontaneously cracks and crumbles, eventually
becoming a loose mass breakable in the fingers, while the
original lump could be broken only with a hammer. In the
cold, therefore, phase 2 is unstable and passes back spontan-
eously into phase 1. Water is set free, probably under great
pressure, and hence the gradual crumbling of the mass. When
compared with the disintegration of minerals, we have here an
example of an enormously rapid chemical reaction, in solids.
The breakage of capillary tubes in the lapse of time instanced
above, § 2, is thus explained.

*The rapidity with which water attacks glass vessels at 210° makes it impossi-

ble to obtain more than an estimate of its compressibility. See my earlier
paper.

Brown University, Providence, R. I.

4 Davison—Platinum and Iridium in Meteoric Iron.

Arr, Il.—Platinum and Iridium in Meteoric Iron; by
JoHN M. Davison.

[Read before the Rochester Academy of Science on October 11, 1898.]

SOME time ago a color reaction, given by potassium iodide
with a residue from a solution of Coahuila meteoric iron, led
me to suspect the presence of platinum. To test the matter,
three attempts were made to separate this element from
meteoric iron; the first with 1086 grams of the Coahuila
iron; the second with 500 grams of the same iron; the
third with 464 grams of the Toluca iron. The hydro-
chloric acid solution of these irons left a fine black sediment
consisting mainly of minute, tetragonal, prisms of rhabdite;
minute, black, irregular crystals, which may also be rhabdite ;
carbon, and a little stony matter.

From this sediment platinum was obtained in each analysis.
There was got from the 608°6 gms. of Coahuila iron 0:014 oms.
of metallic platinum, and 0:0015 gms. of a black powder,
insoluble in nitro-hydrochloric acid, but, after fusion with zine,
dissolving in that acid and giving with ammonium chloride a
dark-red crystalline precipitate, which is probably ammonium
iridi-chloride.

From the 464 gms. of Toluca iron a few crystals of potassium
platinichloride were obtained. These show a reddish color,
and probably contain iridium. The hydrochloric acid treat-
ment of 500 gms. of Coahuila iron left 9°386 gms. of residue:
that of the 464 gms. of Toluca iron left 0°944 gms. of residue,
which contained more stony matter than did that from the
Coahuila. This stony matter was much decomposed by the
acid.

Careful search was made for microscopic diamonds. None
were found in either meteorite. The few transparent grains,
left with the carbon after all soluble matter had been removed,
were decomposed by hydrofluoric acid.

From these analyses all platinum utensils were, of course,
carefully excluded, the reagents tested, and all precautions
taken against accidental contamination.

T. Holm—Studies in the Cyperacec. 5

Art. III.—Studies in the Cyperacee; by THro. Hom.
VIII. On the anatomy of some North American species of
Scleria. With figures in the text, drawn by author.

HAVING discussed the composition of the inflorescence in a
previous article,* let us pass to examine the anatomical struc-
ture of some of our native species: S. pauciflora Mubl., S.
triglomerata Michx., S. oligantha Michx., S. Baldwini Steud.,
S. ciliata Michx., S. filiformis Swtz., S. Hlliotta Chapm., S.
Torreyana Walp., S. reticularis Michx., S. hirtella Swtz. and
S. verticilata Muhl.

The root.

All our species are perennial and the roots are generally very
strong and develop especially from the lower side of the
rhizome ; they possess much the same structure as has been
described before as characteristic of roots of other genera of
this large order, but differ from these, however, by the develop-
ment of the innermost three or four layers of the bark-paren-
chyma into a sheath of very thick-walled cells, which surrounds
the endodermis. This sheath, in connection with the very
thick-walled endodermis, forms a very strong mechanical sup-
port to the central-cylinder. Otherwise the root shows an
epidermis of usual structure, a hypoderm and an outer bark
with large lacunes from the tangential collapsing of the cells,
in contrast to the inner layers (B
in fig. 1), which, as described
above, forms a very solid tissue.

The endodermis represents a
typical QO-endodermis, and _sur-
rounds a pericambium of normal

structure. There are several large ieee
vessels in alternation with groups
of leptome, and the innermost bi nad

-part of the central cylinder is Ceseu wags. party 2

occupied by a large mass of thick- y¢.1. Root of S pauciflora;

walled conjunctive tissue. transverse section. B, inner bark ;
This structure of the root was 224. endodermis; P, pericambium.

observed in species from various ee

localities; from low, swampy ground and from dry, sandy soil

in High Pine woods of subtropical Florida.

* This Journal, vol. v, January, 1898, p. 47.

6 T. Holm—Studies in the Cyperacee.

The rhizome.

While the ramification of the stem underground is sympodial
in our species, we may, nevertheless, distinguish three forms
or modes of growth, viz: cespitose, creeping and tuberous.
The first of these is exhibited by S. vertzeellata, and partly,
also, by S. reticularis and S. Torreyana, and does not differ
from the usual cespitose rhizomes of other Cyperaceee. S.
ciliata and S. hirtella have creeping rhizomes reminding one
of Huirena scirpoidea, which we have described in a previous
paper.* The third form, the tuberous, is well exemplified b
S. pauciflora and partly, also, by S. fileformis and S. Hlliotize.
In these species we notice that the rhizome is very much
branched with relatively short internodes, densely covered by
rudimentary sheathing leaves. The two basal internodes of
each flower-bearing shoot are, furthermore, distinctly swollen
so as to give the whole rhizome a peculiar knotted appearance.
It is the basal internodes of the axis of first order that show
such swelling, while the secondary axis grows out as a short,
creeping and rather slender branch.

The development of these forms of rhizomes does not seem
to depend upon the character of the soil, since the creeping is
noticed in damp (S. hertella) as well as in very dry sandy soil
(S. celiata). The cespitose form is especially characteristic of
S. verticillata, which was collected in damp soil, like S. reticu-
laris and S. Torreyana. The third form is common to species
from dry, sandy to rich or swampy ground. Examining the
internal structure of the rhizome of S. pauciflora, we notice a
thick-walled hypoderm of about three layers inside the epi-
dermis, and bordering on a large bark-parenchyma of thin-
walled cells, which contain deposits of starch. The endodermis
is not very well differentiated, but seems to consist of rather
irregular, thick-walled cells. The innermost part of the rhi-
zome consists of a large, starch-bearing fundamental tissue in
which a number of collateral and perihadromatic mestome-
bundles are imbedded, each of which is surrounded by layers
of stereome. A corresponding structure is, also, noticeable in
the tuberous base of the stems, but in these we observed a dis- -
tinct and thin-walled endodermis, and noticed that the stereome
around the mestome-bundles is only weakly developed.

In WS. ciliata and S. hirtella we find a similar structure, but
the cells of endodermis are more thin-walled. The stereome
when compared with that of S. pauciflora is much heavier in
S. hirtella, and much weaker in S. czlzata. Reservoirs of tan-
nin were observed in all three species, and seemed to be espe-
cially abundant in the bark and fundamental tissue of S. czlzata.

*This Journal, vol. iv, July, 1897, p. 25.

T. Holm—Studies in the Cyperacee. ?

The aerial stem

which bears the inflorescence is sharply triangular in our
species, glabrous or sometimes covered with a few, unicellular
hairs along the upper part of the edges, and these are sharply
pointed and directed upwards. As in most of the other Cyper-
acee with sympodial shoots, the flower-bearing stem is in
Scleria preceded by a few leaves, which winter over.

In its anatomical structure the stem of S. pauciflora shows
a distinct cuticle and an epidermis with very thick outer walls.
The characteristic siliceous cones which we have mentioned in
our previous papers are, also, noticeable in this genus, in those
cells of epidermis which lie over the stereome. The _ bark-
parenchyma contains chlorophyll and consists of a compact
tissue with several tannin-reservoirs scattered here and there.
It is interrupted by the stereome, which covers the leptome-
side of the mestome-bundles, reaching through the bark out-
wards to the epidermis. Viewed in transverse section, the
stereome is seen to form a large group between the sides of
each of the three angles of the stem, while the angle itself is
occupied by bark-parenchyma. The mestome-bundles are
arranged in two alternating bands, the outer consisting of large
and small ones, the inner only of large ones completely im-
bedded in the pith, which is solid in this species. A like strue-
ture occurs in the lateral peduncles, which, however, are cov-
ered by hairs of the same shape which we have described as
characteristic of the main stem.

The structure of the angle of the main stem seems to offer
very good anatomical characters for the species of Scleria, and
the following differences have been noticed. In S. pauciflora
the angle is occupied by bark-parenchyma, which partly sur-
rounds a very small mestome-bundle with little or no support
of mechanical tissue on its leptome-side. Inside of the bark
follow, then, heavy layers of stereome, which form a solid
bridge between the two sides of the angle, and which support
a very large mestome-bundle. This same structure is, also,
characteristic of S. triglomerata, S. oligantha, S. Baldwini
and S. Hilzottw. In S. reticularis, S. Torreyana and S. ciliata,
on the other hand, the angle itself is occupied by a small group
_of stereome which borders inwardly on some strata of bark
parenchyma in whicha very small mestome-bundle is situated.
Next to that occurs a large mass of stereome which connects
the two sides of the angle as in S. pauciflora, and which sup-
ports a mestome-bundle with a hadrome of considerable devel-
opment. WS. verticillata, S. hirtella and S. filiformis illustrate
a third form of structure, since the angle contains here a
single, large mestome-bundle, supported by a strongly devel-
oped stereome on the leptome-side.

8 T. Holm—Studies in the Cyperacee.

Some other divergencies may be observed in the relative
development of the bark-parenchyma and the stereome of the
stem; we notice, for instance, a very thin bark-parenchyma in
S. oligantha, in contrast to S. filiformis and S. cliata, where
the same tissue forms strata of rather large dimensions. The
stereome seems to be present in larger groups and to be more
thick-walled in S. pauciflora, S. Baldwini and S. ciliata than
in any of the other species. The pith is broken by lacunes in
S. Torreyana, S. reticularis and S. filiformis, but is solid in
the other species.

The leaf.

The leaves of our species are provided with a tubular sheath
and a long linear blade, which is very narrow in the North
American species with the exception of S. triglomerata. Scleria
is marked by the total absence of a ligula. Projections from
the epidermis, such as spines or warts, are scarce, and S. ¢rd-
glomerata seems to be the only one of our species in which
continuous rows of rather sharp and curved projections are
visible above the stronger ribs in the leaf-blade. Hairs are, on
the other hand, common, and occur as a more or less dense
pubescence upon the sheaths of all our species, or as scattered,
soft hairs along the margins of the leaf-blade, besides and not
infrequently upon the upper face of the blade, covering the
furrows between the stronger ribs.

The anatomical structure of the leaf-blade seems to be very
uniform in our species, yet exhibits some peculiarities which
enable us to separate the species, at least in groups by anatom-
ical characters. In Scleria paucifiora the stomata are confined
to the lower face of the leaf-blade, forming longitudinal rows
underneath the mesophyll. The cells of epidermis are rather

Fig. 2. Leaf of S. pauciflora; transverse section; BC, the bulliform cells.
x 120.

large and thin-walled on both faces of the blade, and form a
single group of the characteristic bulliform cells above the
midrib (BC in fig. 2). The other cells of epidermis of the
upper face are remarkably large, but without attaining the

T. Holm—Studies in the Cyperacea. 9

shape or arrangement of proper bulliform cells. The meso-
phyll forms throughout a compact tissue, containing much
chlorophyll and is differentiated as a palissade tissue on the
upper face of the blade. Tannin-reservoirs are not uncommon
in the mesophyll. Mestome-bundles of two degrees of devel-
opment separate the mesophyll, which shows a distinct constric-
tion at each bundle, as shown in the accompanying figure 2.
A usual colorless parenchyma-sheath and a rather thin-walled
mestome-sheath surround the bundles, the larger of which are
supported by stereome on both faces, above and below. The
smaller mestome-bundles show only a few stereome-cells on
their leptome-side.

The same distribution of bulliform cells as representing only
one group above the midrib is also met with in other species of
Seleria, and appears to be the commonest arrangement. We
have observed it in S. trzglomerata, S. oligantha, S. ciliata, S.
jiliformis, 8. Elliott, S. Baldwini and S. reticularis. But
in none of these species does the leaf-blade show the deep fur-
rows which we observed in S. pauciflora, and which gives the
section the peculiar constricted appearance. The epidermis-
cells of the upper face are, however, also very large in these

species, increasing in size towards one of the largest of the
lateral ribs, but without becoming specialized as proper bulli-
form cells. It is, furthermore, characteristic of these species
that one or two of the lateral ribs project very strongly on the ~
upper face, while in S. pauczfiora they were all situated below
the surface, in shallow furrows. Stomata occur on both faces
of the blade in these species, in contrast to S. paucifiora. Some,
but slight, variations were also noticed in the thickness of the
cell-walls of epidermis and stereome; we observed, for instance,
a very thick-walled epidermis in S. ciliata, S. Baldwini, S.
juliformis, and S. Hlliottii ; a thick-walled stereome in WS. cidz-
ata and S. Baldwini, besides that the mestome-sheath showed
@ more prominent thickening i in these two species than in any of
the others. S. retecularis is the only species in which the stereome
on the upper face of the leaf-blade is separated by mesophyll from
the mestome-bundles. Characteristic of the remaining three
species: 8. Torreyana, WS. hirtella and SW. verticillata is the
presence of more than one group of bulliform cells, there being
three between the larger lateral ribs in S. 7" orreyand besides
the one above the midrib. In the two other species the epi-
dermis between each two mestome-bundles and above the mid-
rib is developed as bulliform cells, the number of groups
corresponding to the width of the leaves, about four on each
side of the keel of S. vertecillata, eight or nine in S. hirtella.
The other tissues, the mesophyll, with reservoirs of tannin,
the stereome and the mestome-bundles, show much the same

10 T. Holm—Studies in the Cyperacec.

structure as described above for S. pauciflora, as regards their
development and distribution.

Fic. 3. Leafof S. verticillata ; transverse section; BO, bulliform cells. x 120.

Scleria verticillata and S. hirtella differ, however, from the
other species of Scleria, which we have examined, and as a
matter of fact from all the other Cyperaceze which we have
studied anatomically, by possessing some very peculiar incrus-
tations of silica upon the radial walls of the bulliform cells.
The accompanying figure 4 illustrates two bulliform cells in

Fic. 4. Two bulliform cells from the leaf of S. verticillata with incrustations ;
transverse section. x 320.

Fig. 5. A macerated bulliform e¢ell of S. verticillata with incrustations. x 320.

Fig. 6. Incrustations from bulliform cells of S. verticillata. x 320.

transverse section, with incrustations upon the radial-walls, the
outline of which are indicated by dotted lines in our figure.
These incrustations form granular masses of a whitish color
and are present in almost ali the bulliform cells, but varying
somewhat in shape and size. By burning sections of the leat
to white ashes and then adding a few drops of hydrochloric
acid these incrustations become very conspicuous and assume
a clear brown color like sepia. Figure 5 illustrates a macerated
bulliform cell, and we notice here that the incrustation has
taken place to the same extent on both of the adjoining radial
cell-walls. A part of the radial cell-wall upon which the
-incrustation takes place contains also silica and remains as frag-
ments attached to the incrustations after they have been mace-
rated, which is to be seen in the accompanying figure 6. Figs.

T. Holm—Studies in the Cyperacee. ul

a, b and c show the incrustations as seen from within the cell
with a part of the cell-wall still remaining: figure d represents
one of the smallest incrustations, seen from above. These
silicious incrustations dissolved in a concentrated aqueous solu-
tion of potassium in about forty-eight hours.

Silicious incrustations are also known to occur in the Gram-
anew, and Dr. Aug. Grob has lately published a very important
treatise of this subject.* He has, however, confined himself
to the Graminee, and judging from his descriptions and figures,
the incrustations which we have observed in Sclerza differ
very much from any of those that have been found in the
Graminee or other monocotyledonous orders.

The pericarp of Sclerta contains also a vast amount of silica.
Tt lies in two distinct strata of which the outer one is built up
by a homogeneous tissue of polyédric cells; the inner part |
consists of a single layer of cells with very long and thin
radial walls. It is the outer tissue that undergoes the incrusta-
tion, and when macerated the silicified cells form some very
beautiful structures like stars, for instance, in S. pauciflora and
others. Considered from an anatomical point of view, the
North American species of Sclerza show several characters in
common; thus it does not seem probable to distinguish all the
species by means of their structure alone. We have, never-
theless, enumerated the most salient points in the accompany-
ing table, from which the anatomical characters may be readily
seen, at least in some of our species.

LEAF. STEM.
a eee a i aE Se pay. fan aN
© |2 1s8 22/4 [86 1% | Sale.
Pleas: joao og |e 7 ae) ae
2 Si) |S. 2 |S omumene: tara = Ss} os
= an a lau = 4 = Os |
Fo \2s = 4p = Se Ss Bs
> Oe ° o - soe) > SQ | +a
See ae eee he | ee |
© {ss je" lag |- [BS | Pa | be =
~ = do H = |
moo aS |e le lea alee, te.) aS ha S| 2
= aS i|oo [ROO SR los | 2S / 23) 22 S
2 ion Owe eee S 22 =) S| as eg 2
a 2 oc: 2. = 2 9) = ES | 5 = . Lard
B§s| 8s FESS 4s! Sa /2e fos/52/ss| 5] 2
Ba | oS |oa7|seg Ba Bo_: igo e tos lier =
a5 | ss 4 os3° =.2 S25 OF) O45 | Sg | 2 =
ae | = SH2glsedq] Bas jals| ws | | "Sx a a
BF BS BRS BSS BS |2ia ae a8 ccm = =
Am | a
= ale as |
Se palciloras 42 Le. +)—|— NY ort |S Pelee Fe Bla pes
S. triglomerata -.-.-_-- 29) a ee | ee ee
Seolisantha: 2. 5.52. 2-. a. | ee eee ere ae Pe de
Se bald winl 202 22.225. PL fa ae rm A nee | il eo | ea (He
PrecUWIaias 2. 2 eee oak 18 |S ee ee a ee en eee
BMUIOPONS os ae Mf ag | | gE a ee EES ie
Spi a2 + | | oe aa am a ESR
. Torreyana __-------- Se) a re ee.
S. reticularis .-.-=-.. = +{/—{/—;/—/]—;{;—/]—/| + eee dee
ye MUP Ceti pe nts —ft— + — + — | — | — ie + aad
S. vertioulatanc. 252.352 Se) BE ee aman Peg eee Let he

*Grob, August, Beitrage zur Anatomie der Gramineen-blatter, Bibliotheca
Botanica, Stuttgart, 1896, p. 34.

12 T. Holm—Studies in the Cyperacee.

Although the first four of these species are not to be sepa-
rated by the anatomical characters, which we have pointed out,
they are nevertheless readily distinguished by their morpho-
logical; the other species seem on the other hand to be
equally well distinguishable by anatomical and morphological
divergences.

In considering the character of soil and climatological con-
ditions of the localities from where our material has been
brought together, one might feel inclined to designate these
plants as belonging to xero- or hydro-philous societies. Their
structural peculiarities, however, do not point towards any dis-
tinction of that kind. Sclerza paucifiora, for instance, from
wet ground in the vicinity of Washington, D.C., does not
differ in any very marked degree from S. triglomerata, which
_ we collected in dry sandy soil in subtropical Florida; further-

more S. celvata, an inhabitant of the High Pine-woods of
Florida, shows no strictly xerophytic characters when compared
with S. retzcularis, a bog plant. While certain species, as S.
reticularis, S. filuformis and S. Torreyana, collected in a
rather wet ground, possess a stem with a more or less broken
pith, other species from similarly damp localities have perfectly
solid stems. It seems as if the North American species of
Sclerva are very little susceptible to variation due to changes
in environment and that they are not to be placed into any
special society, either among the “‘xero- or hydro-phytes.”

Brookland, D. C., August, 1898.

Starkweather—Regnault’s Calorie, ete. 13

Art. 1V.—Concerning Regnault’s Calorie and Our Knowl-
edge of the Specific Volumes of Steam; by G. P. Stark-
WEATHER.

THE volumes of saturated steam at various pressures custo-
marily used are those determined from the latent heats. Since
the latter depend almost entirely upon Regnault’s experiments,
the question of his ‘calorie’ is of importance. In calculating
the ‘heat of the liquid’ and ‘total heat’ Regnault assumed
that within the calorimeter range, say from 0° to 30° C., the
specific heat of water was sensibly constant. Whether this is
so or not depends upon the thermometric scale used. If mer-
eurial it might be so, but if the scale was that of the air-
thermometer, which is the one to which Regnault usually
reduced his observations, it certainly would not.

Bosscha* in 1874 remarked that in the experiments under
consideration there was no record of the calorimeter tempera-
tures being reduced to the air-thermometer, and accordingly
made corrections to Regnault’s formula for ‘heat of the liquid,’
although he still assumed that the specific heat of water was
constant within the calorimeter range. To this Regnault
replied that the calorimeter temperatures were reduced to the
air-thermometer, but as this was over twenty years after the
experiments were conducted there has always been more or
less uncertainty to many minds.

The writer hopes to present some evidence from Regnault’s
experiments themselves corroborating his statement, and will
construct formule taking into account the variation of the
specific heat of water from 0° to 30°. The agreement between
Rowland and Griffithst regarding this variation is so exact,
and their methods were so accurate, that between the limits
mentioned it can be regarded as determined. Their results
are expressed according to the Paris nitrogen scale, but the
specific heats according to Regnault’s air-thermometer would
not differ qualitatively from them, and ought not to differ
much quantitatively. The only person, so far as the writer is
aware, who has looked at the subject in the light of recent
experiments is Ekholm,t but he concludes (it seems to the
writer incorrectly) that Regnault’s calorimeter temperatures
were not reduced to the air-thermometer, and accordingly
makes no corrections for the variation in the specific heat at
low temperatures, supposing it to be constant according to the
mercurial thermometer.

* Ann. Phys. und Chem., Jubelband, p. 549.

+ Phil. Mag., xl, p. 449, 1895, and Proc. Roy. Soc., lxi, p. 479, 1897.
¢ Bihang till Handlingar Svensk. Vet. Akad., xv, 1889, No. 6, p. 27.

14 Starkweather—Regnault’s Calorie and Our

Before making use of Regnault’s experiments on heat of the
liquid it is well to note the error discovered by Velten* in
them. In Regnault’s table the second column gives the weight
of cold water in the calorimeter, P ; the third, the weight of
hot water, p; the seventh, the temperature, T, of the hot
water according to the air-thermometer ; the eighth, the initial
temperature of the calorimeter, ¢,; the ninth, the final temper-
ature of the calorimeter, ¢,; the eleventh (¢,—z,) corrected ;
the last, the mean specific heat between ¢, and T, z. Then we
should have, according to Regnault’s assumption that at low
temperatures the specific heat is practically constant,

REE PGi)
°= Baad)

Now Velten has found that out of Regnauit’s forty experi-
ments in only five does his value of x agree with this formula;
in twenty-two cases the difference is insignificant, while in
thirteen cases there is complete disagreement. Velten has.
ascribed this discrepancy to errors in calculation, but it is very
strange that erroneously calculated these values of 2 should.
bear a good agreement with the others and as ‘correctly’ cal-
culated by Velten should not. Sutherlandt has almost con-
clusively shown that the error is one in the copying of Reg-.
nault’s numbers into the tables, namely, that Regnault in these
thirteen experiments, which belong to a different series from
the others, copied a wrong set of numbers for the weight of
hot water, and that his values of a are the correct ones.
Sutherland’s proof is based on considerations of the cubic
capacity of the calorimeter, as these very experiments show an
unusually large amount of hot water. In fact, as the num-.
bers stand, there is indicated a total weight of hot and cold’
water greater than the capacity of the calorimeter.

The writer has made corrections to Regnault’s experiments.
on the heat of the liquid necessitated by the variation of the
specific heat within the calorimeter range, but has found that
Ekholm’s formula fits as well as any. Although this was
formed without the above mentioned corrections, nevertheless
the latter are considerably smaller than the variations in the
experiments themselves. Also in developing the formula
Ekholm ieft out certain experiments which are those the cor-
rections affect the most. The formula is

h = t+ 0:00000929 2? + 0:000000265 @’,

and is to hold only above 100°.
The unit here is the specific heat at 15° according to Reg-.

* Wied. Ann., xxi, p. 45, 1884.
+ Phil. Mag., xxvi, p. 302, 1888.

Knowledge of the Specific Volumes of Steam. 15

nault’s thermometric scale, which will be denoted as Regnault’s
calorie. The formula may be incorrect towards 200° by about
0-3 per cent, but is quite accurate at 100°, since in this vicinity
the corrections are very slight whatever be the thermometric
scale. It gives for the mean specific heat between 0° and 100°
the value 1:00358.

Now if Regnault’s scale was that of the air-thermometer,
this number should agree with the determinations made by
other observers who have used that scale or ones not differing
essentially from it. Such determinations have been made by
Von Muenchausen,* Rowland,t Henrichsen,t Baumgartner,§
Velten| (two series, one by the method of mixtures, the other
with the Bunsen ice-calorimeter), Dieterici,4] Ludin,** Joly,tt
and Reynolds.t{

Von Muenchausen’s work, which was conducted by the
method of mixtures, has been criticised by Rowland on the
ground that the experiments were performed in an open
vessel. An inspection of his tables reveals great discrepancies.
Thus as a mean of row I he finds that the mean specific heat
from 26°-70 to 42°°37 is 1:0033 times that from 20°:07 to
26°-70. This is impossible, for the specific heat reaches a
minimum at 32°, by Rowland’s experiments, and there would
be necessitated a mean specific heat from 32° to 42° of 1:0033,
the specific heat at 15°, ¢,,, being the unit. This mean specific
heat would indicate for c,, the value 1°0102, an increase of 1°4
per cent in ten degrees. Moreover in rows II and III he finds
that the mean specific heat from say 25°-7 to 42°°7 is 1:0037
times that from 17°-3 to 25°-7, an increase over the preceding,
whereas according to Rowland’s data it should be less. His
formula makes ¢,_,,, equal to 1:015 ¢,,.

The writer has made careful comparisons of a similar nature
of the experiments of Baumgartner and Velten, and finds
even worse discrepancies. Baumgartner makes the mean
specific heat from 0° to 100° equal to 1:016 ¢,,, while Velten’s
two series make it respectively 0°9896 and 0:9929.

The experiments of Henrichsen were performed with the
ice-calorimeter. Dieterici, who is an authority on the use of
this instrument, has made a careful comparison of them with
those of Velten carried out by the same method. He says
(reference, p. 448): ‘“ The great differences which the experi-
ments made with the ice-calorimeter show, make it probable
that this instrument is unsuited for this investigation.”” More-

* Wied. Ann., i, p. 592, 1877. + Proc. Amer. Academy, 1879-80, p. 120.
Wied. Ann., viii, p- 83, 1879. § Ibid., viii, p. 648, 1879.
|| Ibid., xxi, p. 31,1884. q Ibid., xxxiii, p. 417, 1888.

** Beiblaetter zu Wied. Ann., xx, p. 764, 1896.
++ Phil. Trans., clxxxvi a, pp. 322, 323, 1895. tt Nature, June 3, 1897.

16 Starkweather—Regnault’s Calorie and Our

over Henrichsen makes the ratio of the mean specific heat
from 0° to 100° to that from 0° to 25° equal to 1°025, a num- ~
ber which is markedly larger than has been obtained by any
otaer investigator.

So far we see that little or no dependence can be placed on
the experiments. Rowland’s work should be more reliable.
His experiments make ¢,,_,,, equal to 1:0024 ¢,,_,,, or reduced
to c,, as a unit, ¢,_,,, 18 1°0049 ¢,, which makes ¢))) equal
to 1:0084 c¢,. He mentions two other determinations and
gives as the mean of all three that ¢,_,,, is equal to 1-0041 ¢,,

The very recent work of Ludin was performed with ereat
care. He gives for ¢,_,,, the value 1:00548 c,,

Joly’s determination of ¢,_,,, was obtained ‘by the use of his
steam-calorimeter, and gives for its value 0:9957 c,,. Concern-
ing the accuracy of this instrument we know nothing, and no
details are given. He himself says he obtained much larger
values in earlier experiments, but he considers this to be more
accurate.

We have now left the determinations of Dieterici and
Reynolds. These were both by obtaining the mechanical
equivalent. Dieterici found by electrical methods (and using
the ice-calorimeter) that c¢,_,,, was equal to 42°436 megalergs.
Griffiths* has pointed out that he made the error of using the
“legal” instead of the “true” ohm, which reduces his value
to 42°33. There is also a constant error} in experiments made
by electrical methods of about one-quarter of one per cent due
to the determination of the electrical standard of electromotive
force, which reduces this to 42°225. This makes the mean
specific heat between 0° and 100° equal to 1:00799 ¢,,. The
experiments of Dieterici were performed with great care, were
sixteen in number, and deviated from the mean at most 0°37
per cent, the probable error being 0°05 per cent. But they
have one fault, that there is lacking a conclusive determination
of the quantity of quicksilver displaced in the ice-calorimeter
by a quantity of heat equal to the mean specific heat from 0°
to 100°. Bunsen has found it to be 15°41 milligrams, Schueller
and Wartha 15-442, and Velten 15-47; Dieterici ‘takes the
second value. Velten’s values being uncertain, as has been
shown, it seems as if less weight should be given to his deter-
mination, which would still further reduce Dieterici’s value for
the mean specific heat from 0° to 100°.

In total opposition to Dieterici stand the recent results of
Reynolds. According to him the mechanical equivalent of the
mean specific heat from 0° to 100° is 41°83 megalergs, which is

* Phil. Mag., xl, p. 446, 1895.

qb; pMiae Sescla sas 449, 1895; Proc. Roy. Soc., lxi, p. 479, 1897, and Johns
Hopkins Univ. Circular, No. 135, June, 1898, p. 54.

Knowledge of the Specific Volumes of Steam. 17

equal to the specific heat at about 20°. His experiments are
as concordant as those of Dieterici.

Let us now compare Regnault with these other investigators.
The table below gives the fractional deviations of ¢,_,,, accord-
ing to them from 1:00358 ce...

(1) Experiments which are unreliable.

Von Muenchausen _ (mixtures) +0:0113
Velten (mixtures) —0:0139
Henrichsen (ice-calorimeter) +0°0214
Velten (ice-calorimeter) —0°0106
Baumgartner (mixtures) +0:0124
Joly (steam-calorimeter) —0:0079
(2) Experiments in which evidences of unreliability are not
present.
Rowland (mixtures) +0°0005
Ludin (mixtures) +0:0018
Dieterici (mech. equiv.) +0°0044
Reynolds (mech. equiv.) —0°0050

The first set are grouped so far as possible into pairs of equal
probability and nearly balance. Of the second set the last two
balance, and the deviations in the first two are very small.
As the numbers stand the positive deviations seem to rather
outweigh the negative; some additional evidence on the nega-
tive side is afforded by the following: Sahulka* has found for
the mechanical equivalent at 58° the value 426°262 kilogram-
meters, or 41°81 megalergs, thus making ¢,, equal to 0°9981 «@,,.
Assuming the specific heat to vary linearly with the tempera-
ture from 35° to 100°, this would make the mean specific heat
from 0° to 100° equal to 0:9992 ¢,,, the fractional deviation
from the value we have assumed being thus —0-00436. This
evidence is not entitled to very much weight, however, because
Sahulka’s experiments deviate from their mean over one-half
of one per cent.

It is thus seen that the assumption that Regnault’s thermo-
metric scale was that of the air-thermometer makes his value of
C,-1 agree well with that obtained by others. It cannot be
more than 0°5 per cent out, and is probably much more accu-
rate than that. Any error in it should be ascribed to the unit
in which it is expressed rather than to inaccuracies in the
experiments themselves.

For temperatures below 100° the writer has formed for his
present purposes the following formula:

h = 1°00449 t—0:'0001904 #?+.0°'000001818 @*.
* Wied. Ann., xli, p. 748, 1890.
Am. Jour. Scl.—FourtH Series, Vou. VII, No. 37.—JANUARY, 1899.
2

18 Starkweather—Regnaults Calorie and Our

Two of the constants were determined so as to make ¢,, equal
to unity and ¢,_,,, equal to 100358, while the remaining con-
stant was adjusted to give a good value to ¢,,; the mimimum
value is reached at 35°. A better agreement might be obtained
by the method of least squares, but since c,_,,, is not too well
known and the remaining data all lie below 40° such nicety
would not be justified. Moreover the only use made of the
formula is to obtain latent heats from total heats, and also the
entropy of dry saturated steam, and the term given by this
formula is very small in comparison with the quantities with
which it is combined.

The formule adopted for the heat of the liquid give us the
following for the entropy of water:

nw = dlog,,T—bT +cT’—d.

510

Below 100°. Above 100°.

los*@ = ae _.. 8°173436 8015698

loop aera = 0°156626 1:250174

log 62 s.5eee |. BG o24 4°229792
Oc es |) SB 280 2490°9

This gives the entropy in kilogrammeters (at Paris) per degree,
the zero of entropy being that of water at 0° C. T is the
temperature absolute, absolute zero being taken as —278°°7 C.

Regnault’s ninth memoir, Memoirs of the Institute of France,
vol. xxi, furnishes us with the greater part of our knowledge
concerning the total and latent heat of steam. Other experi-
ments have been made by Andrews,* Favre and Silbermann,t
Berthelot,t Schall,§ Dieterici,| Hartog and Harker, Griffiths,**
G. Fuchs,t+ A. Svensson,t{t W. Ramsay and D. Marshall,§§ and
Harker.|| All exeepting those by Dieterici, Griffiths and
Svensson are at, or in the close vicinity of, 100°. Those b:
Hartog and Harker and by Harker alone are admitted by the
authors themselves to be inaccurate. In all the remainder
there is doubt as to what unit the results are expressed in, and
they are few in number. Taken at their face value they all
give a remarkable confirmation of Regnault’s determination at
100°. The following table, taken partly from Griffiths’ paper
on latent heats, shows this:

* Chem. Soc. Journal, 1849. + Ann. de Chimie et de Phys., xxxvii, 1853.

+t Comptes Rendus, Jxxxv, 1877. § Berichte d. Chem. Gesell., xvii, 1884,

| Wied. Ann., xxxvii, p. 494, 1889. :

4] Proc. Manchester Phil. Soc., 1893-4.

** Phil. Trans., elxxxyi A, p.-261, 1895.

++ Inaug. Dissertation, Erlangen, 1894, and Beiblaetter zu Wied. Ann., xix,
p. 559, 1895.

tt Beiblaetter zu Wied. Ann., xx, p. 559, 1895,

§§ Rep. British Ass., 1895, p. 628.

||| Mem. and Proc. Manchester Lit. and Phil. Soc. (4), x, p. 38, 1896.

Knowledge of the Specific Volumes of Steam. 19

Number of Extreme

Observer. Temperature. Experiments. values of L. Mean L.
LA EEN gr ee 100 8 530°8 -5438°4 =535°9
Favre and Silbermann, 99°81 3 532°59-541°77 = 535°77
ermiclog,2 =... 100 3 535°2 —537°2 536°2
MM en he 100 532°0
DG ere 100 : 536°8
Ramsay and Marshall, 100 537°0
reemault ._.- .__... EEO OS 44 533 0 —538°0 536'3

The greatest deviation of these mean values from Regnault’s is
about eight-tenths of one per cent. To get Regnault’s latent
heat from his total heat I have used the ‘ heat of the liquid’
from the formule already developed.

Consider now Regnault’s experiments on total heats. These
are divided into four classes, those at or near 100°, those from
119° to 195°, those from 63° to.88°, and last the set from —2°
to +16°. He has given as a formula to unite these

H = 606'5+0°305 ¢.

Take the series at 100°, forty-four in number. Neglecting
the first six, which Regnault does not regard as so accurate as
the others, the mean is 636-67, and the greatest deviation from
this is less than three-tenths of one per cent. If we admit the
first six the greatest deviation is less than six-tenths of one per
cent. When we consider the fact that Regnault purposely
varied all his experiments in many ways so far as possible, in
order to eliminate constant errors, this close agreement in so
many experiments shows that the mean value given above is
very. exact.

The unit in which this is expressed is determined by the
calorimeter range, the average of which was from about 8°:3
to 20°°9, from which range the individual experiments did not
differ essentially. Accordingly the unit is the specific heat at
14°°6 according to Regnault’s thermometric scale, or practically
Regnault’s calorie. This we have judged to be the specific
heat at 15° according to Rowland’s scale, and have investigated
the possible error in that judgment. There will be accordingly
the same possible error in the total heats.

Consider next the experiments above 100°, seventy-three in
number. A comparison of the numbers given in Regnault’s
tables with those given by his formula follows: From 119° to
150° the deviations of the experiments from the formula vary
from —0°36 per cent to +0°25 per cent, mostly negative, but
are generally very slight. At 153°, or experiment number 32,
there is a sudden jump to —0°5 per cent, and a deviation exists
varying from —0-2 per cent to —0-°65 per cent, averaging
—0°5 per cent, and gradually increasing up to 175°°5. Here
there is another sudden change, and the deviations vary up to

20 Starkweather—Regnaults Calorie and Our

195° between +0°2 per cent and —0-2 per cent, mostly posi-
tive. These sudden changes are rendered much more promi-
nent by the following table. In the first column is given the
number of the experiment, in the second the temperature, and
in the third the fractional deviation of the observed total heat
from that calculated by Regnault’s formula.

1 119°25 —:0009 —-0014 —-0011|38 1557 —-0040 —-0025 —-0019
11960 —:0018 —‘0018 —:‘0015)39 1565 —-0020 —-0005 +-0003
3 122°17 —v025 —:0028 —-0025|40 1571 —-0046 —-0031 —-0024
4 125° —'0013 — 0003 "0000 | 41 157°8 —-0040 —-0025 —-0018
5 1255 —0018 —-0008 0005 | 42 16073 —:0039 —-0024 —:0016
6 127°2 —-0008 +:0002 0005; 45 1604 —-0034 —-0019 —:0011
7 1290 —0011 —-0001 0002} 44 161°8 —-0030 —-0015 —-0007
8 1344 +°0024 +:0015 © 164.6 —:0050 —-0035 —-0026
9 1342 +°0002 —-*0007 0004 |46 1649 —-0024 —-0009 "0000
10> sock OOS eas 0007; 47 171°6 —-'0050 —:0035 —-0024
11 135°0 +:°0023 +:0014 0017 | 48 1726 —-0059 — 0044 —-0033
12 135°2 —:0002 — 0007.49 1726 —0059 —-0044 —:0033
13. 1355 —'0013 —-0006 "0002; 50 172°8 —-0066 —-0051 —-0040
14
15
16
17
18
19
20

Paco te carl
}
=
(ee)
>
Or

1357 —:0009 —-0002 0002/51 173°1 —-0058 —-0043 —-0031
1364 —-‘0008 —-0001 173-4 —'0059 —-0044 —-0032
1375 —'0016 —-0009 "0005 | 53 1739 —'0056 —-0041 —-0029

} ++ |
So
SS
=)
vo
OL
bo

1377 —0020 —-0013 —-0009|54 1740 —-0055 —-0040 —:0028

1386 —'0006 +:°'000L +°0005|)55 1753 —-0060 —-0045 —-0034

142.00 —0009 —-‘0002 +°0002|56 1755 —-0060 —-0045 —-0034

142°2 —:0016 —-0009 —-0005|57 17973 +°0016 +4°0031 +4°0044
21 142° —'0036 —‘0029 —'0025)58 179°6 +°'0013 +°0028 +:0041
22 1434 +4:0002 +°0009 +°0013|59 179°6 +:0021 +4°70036 +:0049
23 1443 —-0018 —-:0025 —'0021/60 1800 +°0013 +°0028 +:0041
24 1443 —:'0013 —-0006 —:0002)61 183:°2 (0000 +°0015 +°0029
25 1453 +:°:0003 —'0004 +°0001/62 1835 +°0005 +:0020 4-0034
26 145°4 —'0014 —'0021 —'0016)63 183°7 +:°0005 +°0020 +0034
27 145° —-‘0028 —'0035 —-0030/64 183°7 —-0011 +°0004 +4-0018
28 1465 —0002 —‘0009 —-0004|65 186°0 +°0020 +:0035 +:0050
29. 1476 —'0030 —'003T7 —°0032|66 1860 +:0026 +°0041 +:°0059
30 149°0 +°0014 +°0007 +-°0012|/67 187°9 +:0009 +-0024 +0040
31 150°2 +°0005 —'0002 +:°0003/|/68 1882 +4 0026 +°0041 +:°0058
32 15355 —0050 —'0035 —:0029!69 i882 +:°0005 +:°0020 +-0037
33 1541 — 0050 —:0035 —-0028|70 193°8 +:0006 +°0021 +4:0040
34 1551 —:0060 —-0025 —-0019| 71 194°2 —-0022 —-0007 +:°0012
35 155°2 —-0060 —-0045 —-0039|72 194°7 —-0008 +:0007 +:0027
36 1552 —-0028 —-0013 —-0007|73 1948 +:0001 +:0016 +:°0036

37 15533 —'0044 —:0029 —-0023

It is at once evident what a remarkable jump takes place at
number 32. Previous to this the deviations have been both
positive and negative, mostly negative, and have averaged
under 0-2 per cent. Now the deviations are all negative and
rarely under 0-4 per cent. But the reason for this is evident
from a glance at Regnault’s tables. From number 25 to 82
the calorimeter temperatures are very nearly constant, averag-
ing 11°°5 and 238°°5, or the results are in.¢,, asaunit. Now if,
as we have seen to be almost certain, Regnault’s thermometric
scale is that of the air-thermometer, and if we express every-
thing in terms of the specific heat at 15° as a unit, these total
heats should be a little large, since the specific heat at 17° is.

Knowledge of the Specific Volumes of Steam. 21

less than that at 15°. And it is to be noticed that three of the
seven deviations in the experiments cited are positive. From
precisely 32 on, however, the calorimeter measurements have
all very nearly 4°-5 for the lower temperature and 18° for the
upper, and the quantities of heat are therefore expressed in
terms of ¢,. Accordingly the total heats from 32 on are
about 0°2 per cent too small and should be increased by that
amount. ‘This greatly smooths out the break at that place.

This point is also verified by experiments 8 to 23. In 8 to
13 the calorimeter range was 12°°5 to 24°, hence the unit is
€,,, and the total heats should be a little large; such is seen to
be the fact. In 13 to 23 the calorimeter range was 6° to 18°,
the unit accordingly c,,, and the quantities of heat are too
small. We find in fact that all the deviations except one
are negative.

The first seven experiments do not bear this out quite so
well. Experiments 1 to 4 are in about c,, as a unit, experi-
ments -4 to 8 in about ¢,,, and hence the latter should show a
slightly greater negative deviation of perhaps 0:1 per cent.
This is not so, but the experiments are few in number and the
deviations slight.

It should be pointed out here that it is not assumed that
Regnault’s formula is absolutely correct, but it is simply used
as giving a smooth set of values from which to estimate the
concordance of the experiments.

The writer considers that the facts here brought forward
still further substantiate the essential correctness of the com-
parison of Regnault’s calorimeter temperatures with the air-
thermometer. Accordingly he has corrected Regnault’s total
heats so that they should all be expressed in the same unit, ¢,,,
and has given in the fourth column of the table on page 20
the deviations of the total heats so corrected from those yielded
by Regnault’s formula. It will be seen that the formula actu-
ally fits the experiments much better than it appeared, the
greatest actual deviation being about 0°5 percent. This means
in the latent heats a maximum deviation of about 0-7 per cent.

Experiments 1 to 57 indicate that the formula gives values
slightly too great. At 57 there is another peculiar jump of
about 0°75 per cent. The explanation of this is not so evident
as in the case of number 32. It lies in the fact that as the
temperatures became higher the experiments became more
dificult to conduct. Regnault himself says (Memoir, p. 700)
that after number 60 the apparatus required frequent repairs
on account of the high temperatures. It therefore appears
that the highest set of experiments, from 57 on, are of less
weight than the others. If we take the ten experiments com-
mencing with 47 which lie between 170° and 176° we find for
their average fractional deviation —-0043, the maximum being

22 Starkweather—fregnaulls Calorie and Our

—°0051 and the minimum —-0035, the sum of the squares of
the differences from the average being almost nil. If, on the
other hand, we take the remaining seventeen experiments,
lying between 179° and 195°, the average fractional deviation
is +°0022, the maximum +-0041, the minimum —-0007. Here
the extreme differences from the average are twice as great as
in the preceding, while the sum of the squares of the differences
from the average is considerable, thus showing their greater
unreliability, apart. from Regnault’s own statement to that effect.

The writer has therefore considered it proper to change
these average deviations by making the total heat for 190° a
little smaller than that given by Regnault’s formula, The
equation formed is

H = 603-2 + 0°356 ¢ — 000021 2

to hold only above 100°. In the fifth column of nee table
(page 20) there are given the fractional deviations of the cor-
rected total heats from this formula. It will be seen that up
to 26 the experiments are satisfied as closely as possible. From
26 to 57 the deviations gradually increase on the negative side,,
while from 57 to 73 they are still more marked on the positive
side. But 26 to 57 are greater in number than the latter, are
more consistent in themselves, and, as we have seen, are more
reliable. This justifies the greater weight given to them. The
average deviation in the last seventeen experiments is +°0088.
Even if the corrections due to the variation of the specific heat
of water are not allowed, the formula. still represents the
experiments better than Regnault’s.

The formula is correct at 100°, and [ think we can say that
at 200° it cannot be more than 0-4 per cent out, cor responding
to an error in the latent heat of 0°6 per cent.

Consider now Regnault’s experiments below 100°. In the
first six of the twenty-three from 88° to 63° the calorimeter
temperatures averaged about 7° and 19°, hence the quantities
of heat are expressed in ¢,, as a unit. Experiments 7 to 13
have for a calorimeter range 9° to 21°, or are in ¢,, as a unit.
The remaining experiments have the same calorimeter range
as the first six. Taking account of these facts, the fractional
deviations of the experiments from Regnault’s formula are as
given below:

t t t

Less" it +001 9° (81:03 —°004 USS —°008
2 87:83 +001 | 10 80°60 —'005 18 70°49 —'001
3 85°97 —'006 | 11 80°37 —'004 19 69°70 — ‘001
4 85°24 —'005 | 12. 80°17 — ‘001 20 68°01 —°007
5 85°20 000 | 18° 79°55 ‘000 21 66°30 —°002
6 84°88 —'005 | 14 78:28 —'005 22 64°33 —°004
7 83:08 —'005 | 15 76°50 — 001 23 63°02 - 000.
8 82°66 —'OO1l | 16 71°35 —'005 .

Knowledge of the Specific Volumes of Steam. 23

It is immediately seen that Regnault’s formula gives results
uniformly too great. The reason is because the formula is
made to fit his experiments near 0°, which we shall see to be
at fault. The average deviation of the formula from the
experiments is 0-3 per cent.

The experiments near 0° are undoubtedly incorrect. They
were performed by a method different from that which was
used in the remainder, and Regnault himself does not con-
sider them so good as the others, indicating a number of causes
of uncertainty (Memoir, pp. 716-719). Another source of
error lies in the fact that Regnault supposed, in common with
the physicists of his time, that the heat necessary to form
vapor depended only on the final state of the vapor and not at
all on the process by which it was formed. This error did not
enter into the preceding sets of experiments, for in them the
process was always one of sensibly constant pressure. In the
present case, in order to produce the evaporation the pressure
on the liquid was reduced, and the operation was accordingly
not one of constant pressure.

In place of these unreliable experiments we have the recent
accurate work of Dieterici. He found for the latent heat at
0° the value 596°8, the unit being ¢,_,,,. This is the result of
a very careful investigation. He made first a series of eight
experiments, L varying from 595-74 to 597-29. He then made
a better set, three in number, the extremes being 595-98 and
596°63. He still further improved his methods and in two
experiments obtained 596°67 and 597-07. The probable error
is + 0°24 or + 0°04 per cent. In terms of ¢,, as a unit the
latent heat at 0° thus becomes 598°9. Svensson’s determination
is somewhat larger than Dieterici’s, it being 599-92 in terms of
G,-1 28 a unit; it is probably not so correct as Dieterici’s, as
shown by the remarkable concordance in the observations of
the latter. Both have the slight possible error due to the
uncertainty of the constant of the ice-calorimeter, which we
have seen to be about one-quarter of one per cent.

Griffiths’ determinations were of the latent heat between 25°
and 50°, the determinations at 30° and 40°15 being made with
particular care. They are as follows, the unit being ¢,,:

t 25 30 40 40°15 49°8
L 581°9 578°7 572°4 572°6 566°5

These give total heats lying considerably below Regnault’s
linear formula, in this respect agreeing with the experiments
of Dieterici, Svensson, and those of Regnault himself from 60°
to 90°. This strengthens the reliability of the formula we
have formed for the region above 100°, it also lying below the
linear formula. As any expression for the total heat must

24 Starkweather—Legnault’s Calorie and Our

satisfy Regnault’s determination at 100°, this being by far the
most certain of any, the possibility of a linear formula is thus
precluded, although it is evidently nearly such.

If we attempt to form an expression for H to satisfy
Dieterici’s value at 0°, Griffiths’ values from 25° to 50°, and
Regnault’s value at 100°, it will make a positive instead of
negative, will not agree with Regnault’s determinations from
60° to 90°, and will make an abrupt break at 100° where it
joins the formula for the region above 100°. If one neglects
Dieterici’s determination and forms a formula to satisfy Reg-
nault’s value at 100° and Griffiths’ determinations, and to join
smoothly with the formula above 100°, it will give a value at
0° far below Dieterici’s. The writer has therefore formed the
following equation, using Dieterici’s value at 0°, Regnault’s H

: H
at 100°, and making a at 100° equal to that by the formula
used above 100°:
H = 598:9+ 0°442 ¢ — 0:00064 @’.

A comparison of Regnanlt’s experiments with this formula
follows, the fractional deviations being given:

1 +°'0018| 6 —:'0040;11 —°0029 | 16 —:°0031 | 21 +:°0002
2 +°0018| 7 —‘0040;12 +:0001 |} 17 —-0061 | 22 —:0016
3 —‘'0051)| 8 70000 |} 138 +°0011;18 +:°0009 | 23 +:0025
4 —'0041) 9 —:0029|14 —°0037|19 +:0009
5 +:°0009;10 —:'00389|15 +:°0004 | 20 —-0050

The agreement is far better than by Regnault’s formula.
Indeed a much better agreement could not be obtained by any
formula which joins smoothly with one which fits the experi-
ments above 100°. If there is a question as to which set of
experiments to fit best, those above 100° or those below, the
answer is immediately those above, for not only are the
experiments below 100° in less concordance with one another,
as shown by the deviations just tabulated, but Regnault him-
self says that they were more difficult to conduct, the boiling
not taking place steadily, but in great puffs at intervals. The
expression for H will be considered only as provisional, how-
ever; the subject of total and latent heats below 100° will be
discussed again in another paper. The formula does not agree
any too well with Griffiths’ latent heats (adding to them /), the
fractional deviations of the latter from the former being about
—"005:

It should be observed that this formula, as well.as the
formula for total heats above 100°, holds independently of the
judgment made that Regnault’s calorie is equal to the specific

Knowledge of the Specific Volumes of Steam. 25

heat at 15° according to Rowland. For since we know the
value of the mean specific heat from 0° to 100° in terms of
Regnault’s calorie by Regnault’s own experiments, it follows
that 598-9 expresses Dieterici’s determination at 0° in the same
unit as that in which 636°7 expresses the total heat at 100°, be
that unit ¢,, or not. Therefore the writer considers that the
formula expresses the total and latent heats in Regnault’s
calorie with a maximum possible error of 0°5 per cent at the
lower end.

Griffiths believes that the mean specific heat from 0° to
100° is almost exactly equal to ¢,. He bases this conclusion
on the following considerations: He has expressed in terms of
¢,, his latent heats, already referred to, very closely by the
linear formula

L = 596-73 — 0°6010¢.

Extrapolating to 0° and 100° there are obtained 596-73 and
536°63. The first is almost exactly the value of L Dieterici
has found at 0° in terms of ¢,_,,,, the second is what one would
get from Regnault’s experiments on total heat if one supposed
that ¢,_,,, is equal to c,, and that Regnault’s total heat at 100°
is expressed also in terms of ¢,. Thus to support a linear
formula, which type of formula is improbable in itself, he has
to make two suppositions, one that ¢,_,,, is the same as @,,,
which we have seen is not the best value, and the other that
Regnault’s calorie is equal to ¢,,, which we have seen to be
probably true. But these two assumptions are contradictory,
for by Regnault’s own experiments on the specific heat of
water ¢,_,,, 18 1:00358 times his calorie.

From the formule for A and H the writer has obtained the
latent heats and thence the specific volumes of dry saturated
steam. In obtaining the latter there are also required the spe-
cific volumes of liquid water at various temperatures, the rela-
tion between the pressures and temperatures of dry saturated
steam, and the value ofthe mechanical equivalent of heat.
The specific volume of water below 100° is for as many decimal
places as have been used constantly 0-001 cubic meters per
Inlogram. Above 100° Hirn’s* formula has been adopted.
For the relation between pressure and temperature Regnault’s
formula H has been used above 100°, while below 100° his
formula C (with the constants as recalculated by Zeuner) is
taken. These should be sensibly correct except below 40°,
where an error exists due to the fact that Regnault did not

know of the sudden change in ze at 0°. The value of the

mechanical equivalent is taken as 427-03 kilogrammeters (at
* Ann. de Chimie et de Phys. (4), x, p. 32, 1866.

26 Starkweather— Regnault’s Calorie and Our

Paris) at 15° C., according to Rowland. This determination
has been strongly corroborated by recent experiments.* The
volumes follow:

t 0 10 20 30 40 50 60

V 208'°94 107:°88 58505 33°198 19°637 12°0638 7°6703
t 70 80 90 100 110 120 130
V 5°03820 + 3°3942 2°3464 1°6587 1:1973 0°88099 0°65948

t 140 150 160 170 180 190 200
V 0°50152 0°38707 0°30278 0°23982 0:19216 0°15566 012739

Consider the possible errors in these. They may first have
an error of not exceeding one-half of one per cent due to the
determination of Regnault’s calorie in kilogrammeters ; how-
ever much this may be, it is constant for all the V’s. Aside
from this there may be an additional error due to our uncer-
tainty as to the latent heats. This amounts to + 0°6 per cent
at 200°, is practiéally zero at 100°, and is +0°5 per cent at 0°.
Below 40° the V’s are entirely uncertain owing to the errors

there in p and - already mentioned.

In volumetric experiments on steam there is an unusual
amount of difficulty on account of condensation on the walls
of the vessel containing it. ‘This condensation increases very
rapidly as the saturation line is approached, and as a conse-
quence all direct determinations of saturation volumes are far
less reliable than those calculated from the latent heats, for
Regnault so arranged his experiments that this source of error
was entirely avoided. An exception should be made of the
determination of the volume at 0° by Dieterici,t in which
more confidence can be placed, not only from the concordance
of his results, but also from the fact that corrections were
made for condensation.

On superheated steam experiments have been made by Reg-
nault,t Fairbairn and Tate,§ Hirn,| -Horstmann,4 Herwig,**
Meyer,tt+ Battelh,t{ and Ramsay and Young.§$ A large
number of observations prior to Regnault’s are not included
here; these are all at very low pressures, hence large volumes,
for which the substance can be treated as a perfect ¢ gas.

* Johns Hopkins Univ. Circular No. 135, June, 1898, p. 54,
+ Wied. Ann., xxxvili, p. 1, 1889.

+ Mem. of the Inst. of France, vol, xxi, p. 700, 1847.

§ Phil. Trans., clii, p. 591, 1862.

|| Théorie Mécanique de la Chaleur.

§| Liebig’s Annalen, Supplemental Band, vi, p. 51.

** Poog, Ann., cxxxvii, p. 592, 1869.

t+ Chem. Berichte, 1877, p. 2068.

tt Mem. R. Acc. Sc. Torino, (2), xliii, 1893.

§§ Phil. Trans., clxxxiii A, p. 112, 1892.

Knowledge of the Specific Volumes of Steam. 27

For comparison of the experiments the following principles
py
ah
indefinite superheating; for two experiments in which the

have been used : should never exceed its limiting value for

pressures and temperatures do not differ much, should be
almost the same (it would be constant for a perfect gas) ; —
at constant pressure should be greater at greater volumes or
temperatures, at constant volume it should be greater at greater
temperatures or pressures, and at constant temperature it
should be greater at greater volumes or lower pressures. The
pv
T
from the chemical composition of water is that given by Van
Laar,* but since he places the absolute zero at —273°-2 C.
instead of —273°°7, the writer has modified his value to corre-
spond, the number obtained being 3°4673, under the assump-
tion that p is in millimeters of mereury and v in cubic meters
per kilogram.

In this manner the writer has made careful comparisons of
the various experiments. Those of Regnault were all at such
low pressures that the substance should behave as a perfect gas,
which probability is well corroborated by them. The single
determination of Meyer is evidently incorrect, as it gives a
pv
el
perfect gas should be fully two per cent.

The experiments of Fairbairn and Tate have in general
pressures far lower than those of the others, but temperatures

best determination of the limiting value of as calculated

value of of 3472 in a region where the deviation from a

nearly as high; accordingly = should be much greater, whereas

it runs far less. Of their twenty-three experiments six are
comparable by the preceding principles with those of Hirn,
Horstmann, Herwig and Battelli, four more with those of
Hirn, Horstmann and Herwig, four with Herwig and Battelli,
and four either with Herwig or Battelli. In almost every case

the values of = run from two to five per cent less than they

should for any agreement whatsoever, and accordingly there
must have been considerable condensation.
Hirn’s experiments yield great discrepancies among them-
: nang 7
selves. In seven of his twenty-two experiments, in which ca
should steadily increase, the values run
3°376 ye Bo0 4, 866). 31304, 3°337 3°345 3°398
* Zeitschr. fiir Phys. Chem., xi, p. 433, 1893.

28 Starkweather— Regnaults Calorie and Our

Similarly the numbers should increase in each of the follow-
ing columns:

3°347 3°34] 3°350 3°36 1
3°382 3°350 3°382 3°362
3°392 3°286 3°308 3°345
3°362 3°361 3°337

-There are a few other discrepancies, but these involve one-half
of his experiments, and it is difficult to get any concordance by
throwing out a few of the experiments. That there was a
tendency towards condensation is shown by the fact that in

four of his experiments the value a is less than is given by

the saturation line for the same pressure, the difference being
more than can be allowed by any error in the latter.

Suppose we ascribe these discrepancies to condensation,
throwing out accordingly about nine experiments. The
remainder are so few that they are only useful as a corrobora-
tion of others. Three of them are comparable with both
Horstmann and Battelli and deviate from Horstmann about
one per cent, from Battelli about one-half per cent, and this is
in a region where the deviation from a perfect gas is only
about 24 per cent. Five others are comparable with Battelli.
Of these two are in agreement, one deviates about 4 per cent,
and the other two deviate 2 per cent. In the last two the total
deviation from a perfect gas is about 6 per cent.

Of Horstmann’s four experiments one is shown to be incor-
rect by a comparison with the others. The remaining three
agree fairly well with Battelli.

Herwig himself discards fifty-five of his seventy-five experi-
ments as showing condensation due to too near an approach to
the saturation line. Of the remaining twenty, fourteen are in
a region where the substance should behave sensibly as a per-
fect gas. Of the other six, five are comparable with Battelli.
Two are in agreement, while the other three give deviations of
0-6 per cent, 1 per cent, and 2°5 per cent.

We thus see as general conclusions from the foregoing that
among the experiments of Regnault, Fairbairn and Tate, Hirn,
Horstmann, Herwig and Meyer, we have available thirteen by
Hirn, three by Horstmann, and six by Herwig. So many of
Hirn’s and Herwig’s had to be thrown out as to render the
others doubtful, but even accepting them we have found that
in the only comparisons which can be made, those of Horst-
mann and Hirn, there is some disagreement, and the experi-
ments are not sufficiently numerous to give an accurate idea of
the deviation of steam from a perfect gas in various regions.
Nor are they useful as corroboration of Battelli’s experiments,

Knowledge of the Specific Volumes of Steam.

for of the twenty-two only seven agree with his, six are not
comparable, four disagree slightly, and five disagree completely.

We have thus left only the experiments of Battelli and
those of Ramsay and Young. Some at least of the determina-
tions of the former at low pressures are incorrect, for they give

values of - which differ too much from a perfect gas. In

the case of Ramsay and Young we are confronted by a diffi-
culty at the very start. I take the following quotation from
them (reference, p. 122):

“The weight of the water in the tube was ascertained by
determining the products of pressure and volume, altering the
volumes; and this was repeated at different temperatures.
Assuming that if these products for any one temperature were
constant, the density of the steam was constant, viz: the
theoretical density, 9, the weight could be ascertained by the
equation

Vapor density X p. v X 273
11°1636 X 1000 X 760 x (278 + t)°

This expression simplifies to
log W = log p. v+ 4:46179—log (273 +2).

“ During the progress of the experiments it happened that
a trace of water passed up the tube, adding itself to that
already present. This, of course, increased the weight, hence
new measurements were made to determine the amount of the
increase.”

Thus were made the determinations of the weight of water
in the tube, one the mean of experiments where the tempera-
tures were 220° and 230°, the pressures ranging from 2500™™
to 3800, in the second the temperatures were 230°, 240° and
250°, the pressures ranging from 2600 to 4000, and in the third
the temperatures were 250°, 260°, and 270°, the pressures
ranging from 2800 to 4300. Now evidently at such pressures
and volumes (the greatest specific volume came out 0°676) it is
impossible that the steam can be in the condition of a perfect
gas, and accordingly the determination of the weight in each
of the three series is incorrect, and the volumes given by Ram-
say and Young should be diminished. The errors in the three
sets are probably very close to the same amount, for although
the increase in temperature would mean a decrease in the error
due to the assumption that the steam comported itself as a per-
fect gas, nevertheless the specific volumes in the three sets
used for the determination of the weight diminished enough
to offset this.

Assuming then this constant error in Ramsay and Young’s
determinations, let us make a comparison of such of them as

W =

30 Starkweather— Regnault’s Calorie and Our

lie near Battelli’s with his. Below is given a table of the frac-
tional deviation, d, of = according to Battelli from that accord-

ing to Ramsay and Young’; in each case the experiments com-
pared had very nearly the same values of p and 7. Above
each value of d is given a number for reference.

No. | 2 3 4 5 6 i
ad + :009 + °006 +:020 +:°021 —'006. +:003 +:005.
No. 8 9 10 11 12 13 14
d +:010 +027 +:090 —:017 —'°013 —'016 —:015
No. 15 16 Stein! A 18 19 20 Al
d — ‘014 —°007 +:°013 —:'021 —'°019 —:'020 —'017
No. 22 ee 24 25 26 OF 28
ad —'012 —°020 —°'023 —:'030 —°012 0 +°009

Nos. 2, 3, 4 were all at one temperature, about 150° C., while
5 to 11 were at 180°, 11 to 18 at 200°, and the remainder at
2B0C. oe

The sudden changes in the values of din 9 and 10 from
those in 5..... 8, in 16, 17 from those in 11 2. ds omdem
26, 27, 28 from those in 18... . 25, can all be accounted for
by condensation, since Ramsay and Young’s figures, as will be
shown, give marked evidence of condensation as the saturation
line is approached, while those of Battelli do not. But the
remarkable fact is that in the first eight, with the exception of
the fifth, the deviations of Battelli from Ramsay and Young
are all positive, the reverse of what is due to the error in the
latter already mentioned. Suppose we ascribe this to large
condensation in the case of Ramsay and Young. Looking at
11 ....28 we see that the constant error in their experiments
must amount to 1°5 per cent, hence in the first eight there must
have been condensation to the amount of 2°5 per cent.

But there are a few experiments below 100° which can be
compared, and in that region Ramsay and Y oung’s experiments
do not have the constant error already mentioned, the weights
being measured directly. These follow:

Ramsay and Young. Battelli. .
p 4547 39°95 35:3 67°57 85°82 100°27
3 96-06 80°15 33-57 17°851 14°1326 120385
T 348-7 348-4 348-7 35224 352-94 359-24
a 3-415 3-454 3-399 3494 3-423 3°422

Now Ramsay and Young observed that there was far more
condensation below 100° than above, so much so that they
themselves admit that the larger part of their low temperature

Knowledge of the Specific Volumes of Steam. — 331

determinations are worthless. Therefore those given above
should have at least 2°5 per cent condensation ; hence, Battelli’s

determinations being correct, P" by Ramsay and Young should

be, perhaps not 2°5 per cent smaller than by Battelli, for the
latter’s volumes are smaller, but certainly 1°5 per cent smaller.
Instead of that the smallest of Ramsay and Young’s is only
0°8 per cent smaller than the greatest of Battelli’s, while one
of them is larger than any of the latter. We thus see that the
supposition that the deviations in 1....8 are due to conden-
sation on the part of Ramsay and Young involves us in diffi-
culties. At any rate they show conclusively that in the region
considered either Ramsay and Young or Battelli is entirely in
error.

Further, Ramsay and Young’s experiments 11....28 have
all the same weight of water with the exception of Nos. 20 and
22, and hence should have a practically constant deviation to
check with Battelli’s. Nos. 11....15 have such a constant
deviation of about 1:5 per cent, while Nos. 18, 19, 21,
23....26 have a fairly constant deviation of about 2 per
cent, different from the former. Hence we see that the two
investigations are entirely out of accord. One of them is in-
correct, perhaps both.

We will therefore next examine Ramsay and Young’s experi-
ments by themselves. J irst they discard all their experiments
below 75° C. as untrustworthy. At 75° their determinations
show consecutive variations as high as 1°6 per cent. Above
100° C. there was a great deal of condensation in every case as
the saturation line was approached ; this is shown by the rapid
increase of vapor density with increasing pressure. Thus at
140° C. total condensation occurred at a pressure of 2645™™;
the saturation pressure for that temperature is 2717. At 150°,
160°, 180°, 190° and 200° total condensation occurred at pres-
sures of 3442, 4571, 7444, 9337 and 11505™™ respectively ; the
corresponding saturation pressures are 3581, 4651, 7546, 9443
and 11689"™. Aside from this the density showed an impossi-
bly rapid increase before condensation.

There are some other discrepancies in Ramsay and Young’s
observations which will merely be mentioned. Thus if com-
parison is made of those experiments at temperatures 270°,
260° and 250° for which the volumes are given as 0°644, 0°585,

0°529, and 0°435, it will be found that at constant volume —

decreases with increasing temperature, which is impossible.
The experiments below 250° at these same volumes, as well as
pv
5 i

at others, make ~— at constant volume increasing with increas-

32 Starkweather—Regnaults Calorie and Our

ing temperature, which is correct; but the increase is so rapid
as to look alittle suspicious. According to Battelli the inerease
is very slow.

Even if Ramsay and Young’s experiments are substantially
correct far from the saturation line, they cannot be used, for
they contain a constant error, the incorrect determination of
the weight of water present in the tube. There is no agree-
ment with Battelli by which this error can be estimated, nor
can we determine it from our saturation volumes already
obtained, since in Ramsay and Young’s experiments near the
saturation line great condensation occurred.

We have left therefore only Battelli. Ina table below is

° : v : ‘ ‘
given a comparison of i according to such of his experiments

as lie near the saturation line with that according to Regnault’s
experiments, the volumes being calculated from the latent heats.
Each of the latter is given immediately under that one of
Battelli’s which is to be compared with it, and the fractional

deviation of - is given last.

v0 129-14 933078. . 149712," 5 2060-1 3061°9 T971°4 12181°
1h 330°7 352°2 373°33 404°01 417-94 456°6 475-71
a 3°413 3°411 3°392 Biols 3°350 3°276 3°210
p 148°79 35462 760 2030°3 2717-6 7646°4 11689
iy 333°7 353°7 373°7 403°7 413°7 453°7 473°7
= 3°420 3°403 3°373 3°317 3°294 3°196 3°143°
d —°002 +°002 +006 +°017 +°017 +°024 +021

It is at once seen that the discrepancy is very bad. Were
the deviations negative instead of positive they could be
ascribed to condensation in the case of Battelli, and his experi-
ments not near the saturation line would still be available. Let
us now see, supposing Battelli to be correct, where an error in
our saturation volumes can be placed. We have seen that the
greatest error which can possibly be ascribed to the mechanical
equivalent of Regnault’s calorie is 0°5 per cent. This would
make the deviations respectively

—'007 — ‘003 +001 +°012 +°012 +°019 +°016

Further we have seen that the latent heat at 200° C. may be
at most 0°6 per cent too small, although at 100° C. it is correct.
Allowing errors in the latent heats the deviations become

—°007 —'0038 +°001 +°010 +°010 +°014 +°010
The two negative ones, at low pressures, could be ascribed to-

Knowledge of the Specific Volumes of Steam. 33

condensation on Battelli’s part, but the remainder of course
cannot.

| dj
So far we have assumed Regnault’s p and = to be correct,
and we see it is impossible to get any agreement with Battelli.
If we allow an error in Regnault’s p and = we must take them

according to the formula Battelli gives to fit his own experi-
ments on corresponding saturation pressures and temperatures,
for we are going on the supposition that his work is correct.
The change in = changes the saturation volumes, so = 1s
doubly changed. The deviations then become

+°023 + °006 +°018 0 +°007 +°'0138 OL

and things are made worse rather than better.

It is thus a question of whether to discard Regnault’s data
entirely, or Battelli’s. Regnault’s results, with the exception
of total heats at low temperatures, have received essential
corroboration from almost every one who has gone over the
same ground, and this is especially so concerning saturation
pressures and temperatures. On the other hand, Battelli’s
saturation pressures deviate considerably from those of all
other observers, we have seen that in some of his volumetric
experiments at low pressures errors exist, and that finally there
is practically no corroboration between his volumetric experi-
ments and those of other investigators.

The writer therefore concludes that all our knowledge con-
cerning the density of steam is limited to the saturation line,
the experiments on superheated steam presenting discrepancies
which cannot be reconciled.

Am. Jour. Sc1.—Fourts Series, Vou. VII, No. 37.—JAnuaRy, 1899.
3

34 Gooch and Jones—Estimation of Borie Acid.

Art. V.— The Estimation of Boric Acid; by F. A.
GoocH and Louis CLEVELAND JONES.

[Contributions from the Kent Chemical Laboratory of Yale University—LXXVIL.]

THE estimation of boric acid by treating the salts of that
acid with sulphurie acid, distilling with methyl alcohol, evapo-
rating the distillate over magnesium oxide, igniting and weigh-
ing, was proposed by Rosenbladt.* A little later, and without
knowledge of Rosenbladt’s experience, a somewhat similar
process,t which consisted in the treating of the compound of
boric acid with acetic acid or nitric acid, distillation with
methyl alcohol, evaporation of the distillate over calcium oxide,
and ignition of the residue, was described by one of us. In
the course of the development of this process, it transpired
that the insolubility of magnesium oxide retards the absorp-
tion of boric acid by that substance, and that the more soluble
calcium oxide retains boric acid more actively and is therefore
to be preferred.

Points in the treatment upon which special emphasis was
Jaid in the original description of this process were the choice
of a suitable apparatus for the distillation, the employment of
a loosely stoppered receiver for the reception of the distillate
upon slaked lime, the careful removal of water from the sub-
stance in the retort before acidifying and treating with the
methyl! alcohol, regulated use of acid, and care in the evapora-
tion and ignition.

The attainment of good results in this process depends upon
attention to details. Modifications have been suggested by
several investigators. Thus, instead of igniting the calcium
oxide in a large platinum crucible, transferring it to the
receiver to hold the boric acid, and returning the calcium oxide
with the distillate to the same crucible for subsequent ignition
of the residue, as was originally proposed, Penfieldt prefers
to ignite the calcium oxide in a small crucible, to collect the
distillate iz ammoniacal water, to evaporate the latter over the
ealcium oxide in a large platinum dish, and to transfer this
residue back to the small crucible for the final evaporation and
ignition. Kraut§$ suggests a modification of form in the appa-
ratus with no other essential change in conditions. Moissan|
has suggested changes in the apparatus and avoids a transfer of
the calcium oxide—collecting the distillate by itself in a closed
receiver, trapped with an ammonia bulb to prevent the escape
of the boric acid from the distillate. Furthermore, Moissan’s
process calls for the use of an amount of calcium oxide from
fifteen to twenty times greater than that theoretically required.

* Zeitsehr. fiir Anal. Chem., xxvi, 21. + Am. Chem. Jour., ix, 23.
+ This Journal, xxxiv, 222. § Zeitschr. fir Anal. Chem., xxxvi, 3.
|| Compt. Rend., exvi, 1084,

Gooch and Jones—Estimation of Borie Acid. 35

From our experience it seems obvious that the demand for
this amount of calcium oxide arises from an excessive use of
nitric acid in the retort and the consequent modification of
conditions in the distillate. Fortunately this difficulty may be
avoided by the use of a little phenolphthalein as an indicator
in the retort and care to limit the addition of nitric acid to the
amount required to produce distinct acidity. The addition of
a drop of the acid and another of the indicator should be
repeated once or twice during the distillation to insure the
replacement of the acid volatilized from the salt slightly
decomposed in the process. The effect of much nitric acid is
bad, not only because it neutralizes the calcium oxide when it
passes to the distillate, but because when it is used a tendency
is developed on the part of the dried mixture of calcium
hydroxide and borate to puff explosively if this ignition is
begun as soon as the residue is dry. If the residue is heated
gradually and as strongly as possible over a radiator before the
flame is actually applied to the crucible, no such action takes
place; we are disposed to attribute it to the effect of the
nitrate and nitrite, produced by the absorption of nitrous
fumes in the lime, upon the alcohol or other organic matter
retained by the lime in the evaporation and drying unless the
latter process is prolonged at high temperature.

That good results may be obtained with small amounts of
calcium oxide provided care as to the use of nitric acid and the
conditions of ignition be taken, is shown by the figures of the
original description and by the following experiments, in which
phenolphthalein was employed as an indicator and the residue
heated strongly over the radiator before actual ignition.

CaO taken. B20; taken. B.O; found. Error.

grm. grm. orm. grm.
2°3405 01788 0°1792 —+0°0004
1°7620 0°1790 0°1785 —0°0005
2-57 0°1824 0°1840 +0°0016
2°5656 0°1788 0°1786 — 00002

These results are accurate within reasonable limits. On the
other hand, without care to ignite gradually we have noted
errors of from 0:0030 grm. to 0:0060 grm. in the process other-
wise conducted similarly. Doubtless the use of large amounts
of calcium oxide as suggested by Moissan may serve the pur-
pose of diffusing the explosive mixture through a mass of
inert matter sufficient to prevent violent puffing, but care to
heat over the radiator as strongly as possible before opening
the flame directly to the crucible answers the same end. The
difficulty does not exist when acetic acid is used in place of
nitric acid, though even in this case it is safer to use the
radiator in the first stages of heating, thus avoiding the danger
of mechanical loss by too rapid ignition.

36 Gooch and Jones— Estimation of Borre Acid.

Following are determinations made by this method with the
use of acetic acid.

CaO taken. B.O3 taken. BO; found. Error.
orm. erm. erm. erm.
0:9977 0°2065 0°2062 —0°00038
1°0220 0°2067 0°2070 +0:°0003
Tigshen yi 0°2077 0:2075 —0:'0002
lo LO 0°1791 0°1795 +0:°0004

The results of the preceding table, as well as those of the
investigators mentioned, are a sufficient answer to the criticism
of Reischle,* that acetic acid and nitric acid do not liberate boric
acid in the distillation process so that good results may be
obtained. Moreover, it has been shown by one of ust that
even carbonie¢ acid is strong enough to bring about complete
volatility of boric acid with methyl alcohol.

The use of Calcium Oxide as a Retainer.

Quite recently Thaddeefft has advocated the abandonment
of calcium oxide as an agent for holding boric acid in the
evaporation of alcoholic and aqueous solutions on account of
the hygroscopic nature of the oxide and the consequent diffi-
culty of securing it in definite conditions for weighing, and
proposes, instead of using calcium oxide, to retain and estimate
boric acid in solution by converting it into the form of potas-
sium borofluoride. '

In the final modification of Thaddeeff’s method the proposal
is made to liberate the boric acid from its compounds by sul-
phurie acid, to volatilize it in methyl alcohol with the aid of a
current of dry air, to catch the distillate in potassinm hydrox-
ide, to treat the mixture of hydroxide and borate with hydro-
fluoric acid in excess and evaporate in the steam bath, to digest
the residue of fluoride and borofluoride at normal temperatures
for two hours with 50°™ of a potassium acetate solution (sp. gr.
1:14) and for twelve hours more after adding 100° of ethyl
alcohol (sp. gr. 0°805), to filter on paper, wash the residue with
62-72" of alcohol (sp. gr. 0°805), dry at 100° and weigh as
potassium borofluoride, after which the borofluoride is to be
dissolved in boiling water and tested with calcium chloride for
possible contamination by the presence of a fluoride. Plainly
Thaddeeff’s procedure presents at the outset difficulties ; for
besides the inconvenience of conducting long digestions with
reagents of regulated strength, the difficulty of procuring
hydrofluoric acid free from silica, which if present (as it
usually is, in the so-called chemically pure hydrofluoric acid of
commerce) would be retained in the borofluoride as potassium

* Zeitschr. fir Anal. Chem., xxvi, 512.

+ Jones, this Journal, v, 442.
t Zeitschr. fur Anal. Chem., xxxvi, 568.

Gooch and Jones—Estimation of Boric Acid. 37

fluosilicate, the inaccuracy of the dried paper filter, and the
obvious uncertainty of success in an attempt to wash a mixture
of acid potassium fluoride and potassium borofluoride in potas-
sium acetate and alcohol, so that the one shall be rendered
entirely soluble while the other remains sensibly unaffected,—
besides these objections, there is the theoretical probability that
boric acid must be lost by volatilization during the evaporation
of the solution of the mixed salts in the presence of free hydro-
fluoric acid. This last point was put to the proof by submit-
ting to distillation in a platinum retort a mixture of equal
quantities of borax and potassium hydroxide with an excess of
hydrofluoric acid, collecting the distillate in potassium hydrox-
ide, evaporating it to dryness and testing it for the presence of
boric acid. When this residue from the evaporated distillate
was treated with sulphuric acid and methyl alcohol, the burn-
ing alcohol vapor gave plainly the green flame of boric acid.
Another portion showed clearly the presence of boric acid
when acidulated with hydrochloric acid tested with turmeric
paper. No boric acid could be detected in any of the reagents
used. It is plain, therefore, that boric acid does volatilize
upon the evaporation of a mixture of potassium fluoride and
borofluoride in acid solution. The amount of such loss is dis-
closed in the record of the following experiment. Portions of
a standard solution of boric acid, prepared by dissolving a
known weight of anhydrous boric oxide in a liter of water,
were mixed with a solution of potassium hydroxide (free from
silica and standardized by conversion to the chloride) in the
proportions to form the potassium borofluoride, and an excess
of hydrofluoric acid was added. The mixture was evaporated
and the residue was dried and weighed at 100°, the whole opera-
tion being conducted in platinum.

HKF, KFBF; Krror in Error in
equivalent to B.O3 theoretical KFBF; terms of terms of
KOH taken. taken. weight. found. KFBF3. B2Os.

grm. _grm. etm: |. “ernie erm. erm.
0°3531 0°1582 0°5701 0°55380 —0'0121 —0°0033
0°3192 0°1430 0°5154 0°5100 —0°0054 —0°'0015
0°3192 0°1430 0°5154 0°5030 —0°'0124 —0°'0034
0°3192 0°1430 075154 0°5088 —0°0066 —0°0018

0°3192 0°1430 0°5154 0°5114 —0°'0040 —O'0011

In experiments (1) to (8) the volume of the solution evapo-
rated was about 50°. In experiment (4) this volume was
reduced about one-half before acidifying with hydrofluoric acid,
while in experiment (5) the solution was diluted about one-half
before adding the hydrofluoric acid. It is plain, therefore,
that in this single step of Thaddeeff’s process there 1s a con-
siderable error of deficiency. On the other hand, the errors
for the full process as laid down by Thaddeeff have been in

38 Gooch and Jones— Estimation of Boric Acid.

our experience invariably differences of excess—presumably
because the loss due to volatilization of boric acid has been
overvalanced by the inaccuracy in washing. It is plain that
the process can give true indicat‘ons only by the balancing of
considerable errors.

If we take into consideration, therefore, the inevitable inac-
curacy and inconvenience of Thaddeeff’s proposal, it cannot be
regarded as a desirable substitute for the process according to
which boric acid is absorbed and retained for weighing with
calcium oxide, especially since the difficulties in the way of
getting constant weights of that substance are by no means
insuperable.

Thus the following table shows the series of weights taken
in several experiments in bringing calcium oxide to a constant
weight in a 50™* platinum crucible ignited over a blast lamp,
as well as the weight taken after adding a known amount of
standard boric acid solution to the slaked oxide, evaporating, and
igniting. ‘The results recorded are those of experiments made
on days not moist beyond the average and with the greatest
care to approach the limit of accuracy with which calcium
oxide and the boric acid held thereby can be weighed under
ordinarily favorable conditions. The first weight of calcium
oxide recorded under each experiment was taken after a strong
ignition over the blast lamp for about one-half hour. The suc-
ceeding weights were taken after similar ignition of five minutes.
In all cases the crucible was left to stand a definite period in a
sulphuric acid dessicator, and, after the approximate value had
once been obtained, the weights of the preceding weighing
were replaced on the balance before the crucible was taken
from the dessicator. The average of the weights bracketed is
the weight taken as constant for the calculations.

CaO B.03 CaQ+B.0; Ca0+B,03
taken. taken. taken. found. Error.
grm. erm. grm. orm. erm.
0°9505 1°1590
(1) 0°9493 ee 0°2095 1°1588 1-1591 tee +0:°0003.
0°9493
11319 1°3499
lei iky) 1°3474
(2) L3rs Pets: O:2150 1°3465 1°34 75 +98 +0°0010
i-T3oks 1°3476 .
0°8028 0°9205
(3) 0°8025 Be cs 0°1184 0°9208 0°9206 eke. +0°0002
0°8024 | 0°9206
2°6980
(4) 2°6975 2°9043
0°2078 2°9046 +0:°0002

2°6973 ) 2°6973
2°6973

2°9049 } 2°9048
2°9044

Gooch and Jones—Estimation of Beric Acid. 39

Obviously calcium oxide may be weighed with accuracy,
with or without boric acid; but the fact remains that a less
hygroscopic absorbent—one requiring less care in the handling,
is to be desired.

The use of Sodium Tungstate as a Retainer.

In searching for a suitabie material of less hygroscopicity to
replace calcium oxide as a retainer for boric acid, we have
found that sodium tungstate, fused with a slight excess of
tungstic acid over that contained in the normal tungstate (to
insure its freedom from carbonate), answers this purpose excel-
lently. This substance is definite in weight, not hygroscopic,
soluble in water, and recoverable in its original weight after
evaporation and ignition. To test its value as a retainer for
boric acid, portions of it—4 to 7 grms.—were fused and
weighed in a 50° crucible, the tungstate was dissolved in
water and to it was added a known amount of a standard solu-
tion of boric acid. After diluting, mixing, evaporating, and
fusing the residue, the increase in weight should represent the
boric anhydride held by the tungstate. The results of the
accompanying table show how accurately the boric acid is
retained under these conditions: In experiments (8) to (7) the
tungstate, after its first weighing, was dissolved, transferred to
a larger platinum dish and mixed therein with the boric acid.
After evaporation to a suitable volume this solution of tung-
state and boric acid was transferred to the original crucible for
final evaporation and ignition.

Na.W0,+ WO; B,03 BO; Error
taken. taken. found. in B2Os.
erm. . erm. erm, erm.

6°5416 0°1784 01771 —0°0018
7°31384 0°1786 OVS —0°00138
5°50038 0:0950 0°0952 +0°0002
4°1394 ; 0°:0944 0°0944 0°0000
LOS 1 0:2148 0°2149 +0°0001
4:7744 0:2718 0°2702 —0:'0016
6°6470 0°2503 0°2487 —0°0016

It is plain that though the sodium tungstate does not hold
the boric acid with absolute accuracy the errors are not
unreasonable —0:0008 grm. in the mean. Upon substituting
the tungstate for calcium oxide as a retainer in the distillation
process, the results were likewise highly favorable.

We used by preference the apparatus originally proposed,
excepting that the Erlenmeyer flask used as a receiver is fitted
tightly to the condenser and trapped with water bulbs. The
retort is made very easily from a 150° pipette and has the
special advantage that particles of the residue spattering dur-

40 Gooch and Jones—Kstimation of Boric Acid.

ing distillation are easily washed from the walls of the vessel
by a slight rotary motion of the retort. It was found that
special care should be taken to give the tungstate ample time
for contact with the distillate before exposing the latter to
atmospheric evaporation. The distillate was received, there-
fore, in a dilute solution of sodium tungstate placed in the
receiver, cooled by ice and trapped with water, and the mix-
ture was well stirred, allowed to stand one-half hour, evapo-
rated to small volume in a large dish, and transferred to the
crucible in which the tungstate had been originally weighed.
After thorough drying the residue was ignited to fusion and
weighed. When acetic acid was employed in the retort, care
was taken in the ignition to expose the fused mass freely to
the air (by causing it to flow upon the sides of the crucible)
until the color of the cooled tungstate was white, in order that
the reducing effect of the acetate might be eliminated. In the
experiments recorded in the following table the tungstate was
used in the receiver to retain the boric acid distilled as usual
with methyl alcohol, from the borates treated with acetic acid,
nitric acid or sulphuric acid, in amounts regulated by the use
of phenolphthalein as an indicator.

Na,.W0,+ WO; B.03 B.03
taken. taken. found. Error
grm. erm, erm. erm.
With Nitric Acid.
8°5516 0°1582 0°1572 —0:0010
4°9639 0°1329 0°1323 —0°0006
8:0033 0°1267 0°1256 —0°0011
With Acetic Acid.
4:9658 - 0°1434 0°1418 —0'0016
6°0289 0°1431 0°14383 +0:°0002
4°6797 0°1589 O'1587 —0°0002
4°0013 0:'1433 0°1422 —0:0011
With Sulphuric Acid.
6°3439 0°1582 0:1579 —0:°0003
8°8227 0°1582 0°1577 —0°0005
10°1516 0°1265 0°1264 —0°0001
6°5738 0°1392 0:1390 —0°0002

Excessive use of acid is disadvantageous, and this is espe-
cially true in the case of sulphuric acid; for, if this acid is
carried over with the methyl alcohol, as it is at 100° if present
in appreciable excess, a part of it, at least, is held permanently
by the tungstate to increase the apparent weight of the boric
acid to be estimated. . |

The manipulation of the tungstate presents no difficulties,
and the results obtained by its use are reasonably accurate.

A. E. Verrill—New Actinians. 41

Art. VI.— Descriptions of imperfectly known and new Actin-
tans, with critical notes on other species, Il.; by A. E.
VeERRILL. Brief Contributions to Zoology from the Museum
of Yale College, No. LIX.

Family Hatocuavip# Ver. nov.

CoLuMN much elongated, either smooth or with suckers ;
base with a very small or rudimentary disk; tentacles 20,
obtuse or capitate; ten pairs of mesenteries, all perfect.

Haloclava, gen. nov. Type H. producta(Stimp.) Ver. Figure 7

Column much elongated, soft, without a sheath, the upper
part with rows of distinct adhesive suckers. Upper end with
a sphincter and capable of involution. Aboral end often
inflated or bulbous, with a rudimentary disk, capable of adhe-
sion to small pebbles. Tentacles twenty, short, usually clavate,
nearly equal, marginal. Mesenteries very muscular, all perfect,

Fig. 7.

arranged in 10 pairs; six of these pairs are larger and form a
regular primary cycle; four other pairs, narrower below,
belong to the second cycle; the usual pair, next to the ventral
or sulear pair of directives, is lacking on each side.

Besides the type species this genus will include /. albida
Ver., which I now consider only a variety. H. Capensis Ver.,
Commun. Essex Inst., v, p. 5, pl. 3, fig. 7, 1869 ;* Z. breve-
cornis (Stimp.) Ver. , Op. cit. p- 4, pl. = figs. 22", QP. and /7.
Stumpsoni Ver., op. ‘cit. aT: 5, pl. 3, fic. i 1869, resemble this
genus but probably belong to Eloactis Andres. No specimens

* Andres, in his classical work on the Actinaria, has entirely overlooked my
two illustrated articles on Actinaria of the North Pacific Expl. Expedition, in the
Comm. Essex Inst., vols. v and vi, 1868, 1869, quoted above, and has, therefore,
omitted one of the above named species from his work, as well as numerous
other species and several genera first described and illustrated in those papers.
In several cases he stated that no figures nor descriptions of certain species
therein described had been published. Nevertheless he quotes those papers in
his bibliography. In the papers referred to there are four plates of Actiniz (40
figures). He also failed to quote several other papers by me. Among those
omitted are Report on Invert. of Vineyard Sound, etc., 1874; Radiata from Coast
of North Carolina, this Journal, iii, p. 432, 1872; Bulletin, U. S. Nat. Mus., No.
15, 1879; American Naturalist, ii, p. 251, and several others. He intended to
include all the literature and all known species. Such omissions are liable to
happen in the best works.

492 A. EF. Verrill—New Whee eee:

of either species were preserved by Dr. Stimpson. His origi-
nal drawings show no suckers.

H. producta ranges from S. Carolina to Cape Cod. Var.
albida, Cape Cod.*

Eloactis brevicornis and E. Cupensis are from Cape of Good
Hope.

EF. Stimpsoni, Hong Kong, China.

Family Bunopactip& Verrill = Bunodide Gosse.

Bunodactis V., new name = Bunodes Gosse (restr.)

Bunodes (pars) Gosse, 1855, 1860, not of Eich. (genus of Hurypteroidea), 1854.

Unfortunately the name Lwnodes was preoccupied,t as indi-
cated above, for a genus of eurypteroids, and some other name
has to be chosen for this well known genus of actinians. None
of the numerous groups recently established seem to corre-
spond with this genus, and I have, therefore, given it a new
name.

This genus, as now limited, includes species that have from
12 to 48 pairs of perfect mesenteries,t the number increasing
with age. All the larger mesenteries, except the directives
may bear gonads. The submarginal sphincter is prominent,
endodermal and circumscribed. The- verrucee of the column
are intraseptal evaginations of the wall and serve as true adhe-
sive suckers; some of the rows may reach the base. The wall
is imperforate, at least in the typical forms. The marginal
tubercles are simple and similar to those below, and are not
true acrorhagi, though sometimes pigmented distinctively.
The wall is thick and muscular.

Bunodactis seems to include a considerable number of
species, widely distributed. Among these are the following:
B. verrucosa (Penn.) = gemmacea (type); B. Balla (Cocks) ;
B. Lnsteri (Johns.); B. glandulosa (Otto) = B. rigida (And.) ;
B. thallia (Gosse); B. eee (Andres), all from Euro-
pean seas; 6B. pluvia (Dana); B. crwentata (Dana); B. ocellata
(Less. ), and B. papillosa (Less.), West S. America ; "B. inornata
(Ver.) Hong Kong; and B. Japonica (Ver ) Japan. In
accordance with the usual custom, the change in the typical
generic name will make it necessary to change the family name
Bunodide to Bunodactide. Tealiade (Hert.) might be used,
were not Zealiaa synonym of Urticina (Ehr., 1834, restricted).

* Andres has based a new sp. (H. Elizabethe) on an incorrect outline figure of
this. in a popular work. It is the same as ZH. albida.

+ Kichwald, Bull. Soc. Imp. Nat. de Moscou, vol. xxvii. part I, pp. 50, 107, pl. II,
figs. 2-4, 1854, Bunodes lunula. The article is dated, at the end, Nov. 2,1853.

t Y.and A. F. Dixon (Sci. Proc. Roy. Dublin Acad., ii, p. 310) have shown
that in B. thallia the number of mesenteries and their arrangement are remark-
ably variable, and that abnormal specimens are found with one or with three

pairs of directives and corresponding siphonoglyphs. I have found the same
variations in B. stella, which usually has about 24 pairs of perfect mesenteries.

A. E. Verrill— New Actinians. 43

Bunodactis stella Verrill.

Bunodes stella Verrill, Revis. Polyps E. coast of U.S, p. 16, pl. I, figs. 1-8,
1864; Amer. Naturalist, II, p. 258, 1868.

Bunodes spectabilis Verrill, Check List Mar. Invert. N. New Eng., p. 15, 1879;
Cont. to Nat. Hist. of Arctic America, in Bulletin U. 8. Nat. Mus., No. 15, p. 152,
1879 (non Fabr.).

Bunodes (?) stella Andres, Attinie Golfes von Neapel, p. 230, 1884. MceMurrich,
Actinaria of Bahama Is., p. 30, 1889.

Having received large specimens of this species from Green-
land and Cumberland Gulf, I at one time identified it with A.
spectabilis Fabr., owing mainly to the green color of the latter,
but as this was not described as verrucose, it is more probable
that it was the green variety of Urticina crassicornis, which
also oceurs in Greenland. At Eastport, Me., between tides, a
common variety of the latter is green mottled more or less with
red or brown. |

This species is a true Bunodes, closely related to the common
European species (L. verrucosa), as stated in my detailed
description in 1864. It is somewhat singular that Andres,
with my full description and figures, should have put it as a
doubtful species, even suggesting that it might belong to
Aulactinia, a genus established by myself for a_ species
described in the same paper with the present one, asif I did not
know the characters of my own genera and species !

MeMurrich simply followed Andres in expressing this doubt,
but regretted that he could not obtain examples for study,
after futile efforts. It is, however, a very abundant species at
low-water mark at Eastport, Me., and numerous specimens have
been preserved by me for the Mus. Comp. Zool. (including the
original types); the U. S. Nat. Mus.; the Mus. of Yale Univer-
sity, etc. He certainly did not ask me for specimens.

He also described a very different species, from the
Bahamas, for which he proposed the provisional name of
Aulactinia stelloides, but he, at the same time, expressed the
opinion that it might prove to be identical with B. stella.

Mr. Duerden (Journ. Inst. Jam., ii, p. 454, 1898), misled by
this statement, went farther and “definitely united the same
species, from Jamaica (as Auwlactenza stella), with my B. stella.
The latter is a very arctic littoral species, ranging from Maine
to Greenland. Its occurrence as a littoral species in the West
Indies would, of itself, be a case without parallel. The West
Indian species probably belongs toa different genus, but it
does not agree with the characters of typical Awlactinza.

For such species I propose the name Lunodella.

Bunodella V., gen. nov. Type B. stelloides (McMur.)

Form of body and tentacles same asin Bunodactis. Column
with longitudinal rows of true adhesive verruce on the upper

44 A. F&. Verrill—New Actinians.

part. Margin with arow of simple, round or conical verruce
similar in structure to the others. Mesenteries in several
eycles with but six perfect pairs (in the type). Submarginal
sphincter well developed, endodermal, circumscribed.

Differs from Awlactenza in not having lobate marginal ver-
ruce. The structure of the column-wall is also different:
From Lunodactis it differs, so far as known, in having but six
pairs of perfect mesenteries, as in very young examples of the
latter. Should larger specimens of Bunodella be found to
have a greater number of perfect mesenteries, this group
would not differ essentially from Bunodactis.

Anthopleura Dowii Ver. Figure 8.
Trans. Conn. Acad. Sci., i, p. 474, 1869; Andres, op. cit., p. 223, 1884 (Dovit).

This species was pretty fully described by me in 1869, but
the anatomy was not then given.

A medium sized example, 40™™ across, in alcohol, has about
96 pairs of mesenteries, of which 48 are perfect and fur-
nished with large plicated muscles alone most of their breadth; ~
a pair of very small imperfect mesenteries, without gonads,
usually intervenes between each perfect pair. The mouth has
two large siphonoglyphs and about 22 crenulations on each
side. Sphincter muscle well circumscribed, ovate in section,
not very large. Larger submarginal tubercles or actinobranchs
have 4 to 6 lobules on the lower side, each with a small point,
appearing as if perforated, though the openings, said to exist
in life, are not visible (fig. 8). Adhesive, cup-shaped suckers,
of the longer rows, reach the base. Panama to San Salvador.

Anthopleura Stimpsoni Ver., Comm. Essex Inst., vi, p. 32, 1869.
Figure 9.

This species has 12 pairs of perfect mesenteries, with large
muscles along most of their breadth ; 12 pairs of 3d cycle, well
developed; about 24 pairs of 4th cycle, of small size and
barren; all the others, except directives, bear large gonads.
Sphincter muscle circumscribed, rather large, ovate in section.
Actinobranchs (fig. 9) crowded, three to five in the larger rows,
the upper one three to five lobed or digitate, the larger lobes
often subacute and perforated. Adhesive suckers large, in
unequal rows. Hong Kong, China. Type.

Bunodosoma, gen. nov. Type, B. granulifera (Les.).

General form and appearance as in Bunodactis, but the
hollow verrucse, arranged in vertical rows, are rounded or
subconical and do not form adhesive suckers. Upper or sub-
marginal ones are larger and in mature specimens more or less

A. FE. Verrill—New Actinians. 45

lobulated, but have nearly the same structure as those below,
though they are described as perforated when living. Tentacles
numerous; many mesenteries, 12, 24, or more pairs being
perfect. Sphincter muscle well developed, endodermal and
circumscribed.

To this group I refer the following:

Bunodosoma granulifera (Les., 1817).

Aulactinia granulifera Andres, 1884, p. 221.
Bunodes granulifera Duerden, 1898, p. 454.
Bunodes teniatus Mc Mur., 1889, p. 23.
Cereus Lessont Duch. and Mich., 1860, p. 42.

West Indies.

Bunodosoma cavernata (Bosc, 1801).

Bunodes cavernata Ver., 1864. p. 17.

Cladactis cavernata Ver., 1869, p. 473.

Phymactis cavernata Andres, 1884, p. 231; Me Mur., 1887, p. 51.

South Carolina to Cape Hatteras.

Various other species described under Bunodes, Aulactinia,
etc., may belong here, but the descriptions of the verruce
and other parts are insufficient to determine their position.
Aulactinia crassa Andres looks more like Bunodosoma than
like Aulactenia.

Family PayLiacrip”.
Asteractis Verrill, 1869. Type, A. Bradleyi.

Asteractis Verrill, Trans. Conn. Acad. Sc1., i, p. 465, 1869. Andres, op. cit.,
p. 292, 1884,

This genus is characterized by the presence of a special wide
ruffled collar outside the tentacles, covered with complex rows
of compound verruce or lobed outgrowths,* but not forming
true free fronds. The column bears more or less numerous
adhesive verruce in rows below the collar. There are 12 or
24, or more pairs of perfect mesenteries and one, or sometimes
two, endodermal sphincter muscles.

Asteractis flosculifera (Les.) Ver.

Actinia flosculifera Les., Journ. Acad. Sci. Philad., i, p. 174, 1817 (not Oulactis
Hlosculifera Duch. and Mich., nor of McMurrich),

Evactis (?) flosculifera Andres, op. cit., p. 235.

Oulactis fasciculata McMurrich, Proc. Philad. Acad. Sci., 1889, p. 108; also in
Heilprin’s ‘‘ The Bermuda Islands,” p. 112, pl. x, fig. 5 (section), 1889.

This species is common at Bermuda, living buried to its mar-
gin in shell-sand at low-water mark. Figur es from life will be
published soon in Trans. Conn. Acad. Science.

* They are probably branchial in function and may be called actinobranchs
(Actinobranchie). The groups that they form may be called pseudofronds,

AG: , A. E Verrill—New Actinians.

It has been pretty fully deseribed, as to its anatomy, by
McMurrich. The collar is covered by radial pseudofronds
composed of about three rows each, of
lobed tubercles, the lobules being short
' or rounded; a single larger lobed tuber-
cle or actinobranch occupies the distal
end of each group, and a nearly simple
one often forms the proximal end, next
to the fosse at the bases of the tentacles.
The color is variable, but the collar is
usually umber-brown, varied with yellow-
ish, flake white, and reddish brown spots.

Asteractis Bradleyi Ver. Figures 10, 11, 12.
Trans. Conn. Acad., i, p. 465, 1869, Andres, op. cit., p. 292, 1884.

This is closely allied to the preceding species. The pseudo-
fronds of the collar are made up of more finely lobulated ver-
ruce, so that they appear more complex. These verruce, in

the alcoholic specimen, are closely crowded in each radial
series, of which there are 48; their upper and lateral surfaces
are divided into many small rounded or conical lobules, most
prominent along the outer sides and at the distal end of the
last or distal one, where about five lobules project, in each
group, beyond the edge of the collar, giving it a finely serrate
or fringed appearance (figs. 11, 12). The verruciform adhesive
suckers are in 48 rows, of 3 to 5 each, corresponding to the

A. E. Verrill—New Actinians. 47

pseudofronds, and extending only a short distance below the
collar.

The general figure, now given, is based on a sketch from
life, by F. H. Bradley, of the original type, corrected by com-
parison with the preserved specimen, which was from Panama.

Asteractis conquilega (Duch. and Mich.) Ver,

Oulactis flosculifera Duch. and Mich., Corall. Antill., p. 46, 1860 (non Les.) ;
Supl., p. 129, 1866. McMurrich, Actin. Bahamas, p. 56, pl. I], fig. 2, pl. IV, figs.
12-14 (anatomy), 1889.

Oulactis conquilega Duch. and Mich., op. cit., pp. 49, 89, pl. VII, figs. 7, 11,1860.

Oulactis foliosa Andres, op. cit., p. 290, 1884.

The collar of this West Indian species is very broad and is -
covered with about 24 pseudofronds, or groups of small folds
and lobulated and crowded actinobranchs, much more complex
than in either of the preceding species.

Asteractis formosa (Duch. and Mich.) Ver.

Oulactis formosa D. and M., op. cit., p. 47, pl. VIL, figs. 4, 5, 1860; Supl., p. 36
(130), 1866. Andres, op. cit., p. 291. 5;

This is closely allied to the preceding species, but has smaller
and less complex groups of lobulated actinobranchs on the
collar and fewer tentacles. Guadaloupe.

Family DenpRometip2 McMurrich.
Lebrunia Duch. and Mich., Corall. Ant., p. 48, 1860.

Actinodactylus (pars) JJuch. and Mich., op. cit., p. 44, 1860.

Stauractis (pars) Andres, op. cit., p. 255, 1884 (new name for last).

Taractea Andres, op. cit., p. 284, 1884 (for LZ. Dane).

Lebrunea MecMurrich, Actinaria Bahama Is., p. 33, 1889; Duerden, Actin.
Jamaica, p. 456, 1898.

This remarkable genus is characterized by having a series of
about six (sometimes 5 or 8) large, dichotomously branched
fronds, or branchize (actenobranchie) exterior to the tentacles.
The species examined by me has, on these fronds, at the forks,
many more or less spherical bodies having the structure of
acrorhagi. There are numerous mesenteries, most of them
perfect and bearing gonads. Sphincter muscle feebly devel-
oped or wanting.

The genus Actinodactylus Duch., 1850, was based on A.
Boscit D., which may, perhaps, have been only a young
Lebrunia. It was only 4 to 5 lines in diameter, with 15 small,
simple tentacles, in a single circle, and with 5 elongated, pedi-
celled, trilobed fronds, the lobes with small subdivisions. It
may, however, prove to be a distinct genus when larger exam-
ples can be studied.

48 A. FE. Verrill—New Actinians.

The second species referred to it (1860), as A. neglectus,
appears to be the young of Lebrunza neglecta, described in the
same work. Although it was only 5 lines in diameter it had
30 tentacles and 5 dichotomously branched fronds.

Lebrunia Dane (D. and M.) Verrill. Figure 15.

Oulactis Dance Duch. and Mich., op. cit., p. 47, pl. VII, fig. 10 (frond), 1860.

Rhodactis Dane Duch. and Mich., Supl., p. 37, 1866.

Taractea Dane Andres, op. cit., p. 284, 1884.

Lebrunea neglecta Duerden, Actin. Jamaica, p. 456, 1898.

? Hoplophoria coralligens Duerden, loc. cit. (non Wilson).*

Specimens of large size, up to 8 inches or more in diameter,
were obtained by me at Bermuda, in 1898. These agree very
well with the specimens described by Duerden from Jamaica,
as L. neglecta, but not with those described by MeMurrich
under the same name. — In having numerous rounded acrorhagi
on the actinobranchs or fronds (fic. 15) my species agrees with the
Oulactis Dane D.and M., from St. Thomas, which is evidently
a Lebrunia, and there is no reason to doubt its identity with
the Bermuda species.

My larger examples were usually dark green in life, with
the tentacles somewhat paler green and flecked more or less
with white; disk olive-green with whitish radial spots; fronds
bluish ereen with the tips and rounded bodies light blue. It
lives with the body concealed in holes or crevices of the reef-
rock, at and below low-water mark. The tentacles are long.

Lebrunia neglecta Duch. and Mich.

Lebrunia neglecta Duch. and Mich., op. cit., p. 48, pl. VII, fig. 8, 1860 (young)..
Andres, op. cit., p. 362 (non Duerden).

? Actinodactylus neglectus Duch. and Mich., op. cit., p. 44, 1860 (very young).

? Stauractis incerta Andres. op. cit., p. 255, 1884 (new name for last).

Lebrunea neglecta MeMurrich, Actin. Bahama Is., p. 33, pl. I, fig. 7 (general), pl.
III, figs. 11-14 (anatomy).

This species was originally based on a small specimen (about
one-half an inch high), but it had relatively large, much dichoto-
mously divided fronds, according to the figure. MeMurrich
has given a detailed description of a species that appears to be
the same, though much larger in size. His figures represent
the 6 fronds as dichotomous, but with few divisions, though in

*The Hoplophoria coralligens Wilson, was described from a single, very small
specimen (diameter about 2™™), having 48 unequal tentacles and 4 simple, elon-
gated, marginal fronds. Duerden identified with it a small form of Lebrunia, hay-
ing dichotomous fronds and agreeing nearly in color, etc., with L. Dane. The
identity of his specimens with the true Hoplophoria seems to me very doubtful.
This generic name is unfortunate for its variants, Hoplophorus and Hopliphorus, had
been used previously for five distinct genera. It evidently cannot be the young
of either nominal species of Actinodactylus, for although much smaller, it has more
tentacles. It is most likely to be the young of Diplactis Bermudensis MeMur., or
some similar species.

A. F.. Verrill—New Actinians. 49

his description it is stated that they ‘“‘dichotomize many times
until a dendritic structure is produced.” The rounded bodies
are not figured nor described by him, nor by D. and M. as
occurring on the fronds. In this respect the fronds are very
different from those of the Bermuda species studied by me.
I am, therefore, disposed to believe that there are two large
West Indian species.

Subfamily Anicunz = Family Auiciap# Duerden.

Duerden, Ann. and Mag. Nat. Hist., 6, xv, p. 213, 1895 ; Haddon and Duerden,
Sci. Trans. Roy. Dublin Soc., vi, p. 153, 1896.

This family was established to include certain tuberculated
genera, such as Alicia (= Cladactis Panc.), Cystiactis, ete.

It is characterized by a thin wall; feeble, diffused sphincter
rauscle; hollow tubercles, usually compound, on the column;
12, 24, or more pairs of perfect mesenteries ; no acontia.

The wall-tubercles are thin and doubtless serve as branchie ;
they are not adhesive suckers.

The group is practically equivalent to tuberculated Actenzda, —
and might well form a subfamily of that family.

Hucladactis Ver.,new name. Type, &. grandis V.
Cladactis Ver., Trans. Conn, Acad. Sci., i, p. 472, 1869 (non Panceri, 1868).

By a mere coincidence the name Cladactis was given by M.
Panceri and myself to two distinct genera, at nearly the same
time. The former appears to be the same as Alzcza Johns.
which preceded it. No other name seems to have been given,
as yet, to my Cladactis. The latter is characterized by very
numerous tentacles, which are scarcely retractile; column
thickly covered with hollow clavate and lobulated tubercles,
the upper ones larger and more lobed than the rest; margin
with a deep fosse; sphincter muscle very feeble and diffuse, or
almost lacking; 36, 48, or more pairs of perfect and many
imperfect mesenteries, having feeble muscles extending along
most of their breadth; gonads numerous, borne on most of
the mesenteries, both perfect and imperfect; a distal septal
foramen; mouth large, with two large equal siphonoglyphs;
lips with very numerous lateral furrows.

Hucladactis grandis Ver. Figures 13, 14.

Cladactis grandis Ver., Trans. Conn. Acad. Sci., i, p. 473, 1867. Andres, op.
cit., p. 226, 1884.

The larger specimens of this species are over 2 inches in
diameter and height, as preserved in alcohol. The tentacles

Am. Jour. Sc1.—Fourta Serizes, Vou. VII, No. 37.—January, 1899.
4.

50 A. E. Verrill—New Actinians.

are partially contracted in length and suleated, but are rarely
concealed, even in the young, by the infolding of the margin
of the disk; generally they are fully exposed and the margin
is but little contracted or not at all, owing to the feeble devel-
opment of the circular muscles. There are about 480 tentacles
in the larger specimens; they form five or six crowded rows,
the inner ones standing about half way between the center and
the margin of the disk; the inner ones are larger and some-
what swollen at the bases but are not much longer than the
outer ones; all are sulcated by contraction. The lips have
about 48 to 60 furrows on each side. The mesenteries are
hexamerous and, like the tentacles, vary in number according
to the age. A medium-sized specimen had about 120 pairs, of
which about 48 were perfect, but the arrangement is more or
less irregular, for one, two, or even three imperfect pairs may
intervene between the various perfect pairs. The larger speci-
mens have about 60 pairs perfect and about 180 imperfect.
Nearly all bear gonads. The larger upper actinobranchs (fig.
14 a, c), next the smooth fosse, are elongated, more or less clavate,
the distal portion divided into four to six irregular rounded or
obtuse lobules ; those lower down become shorter and less lobu-
lated, but nearly all are clavate or capitate, to the base, and most
of them, in large specimens, are divided into two, three, or more
lobules. They are thickly crowded over the whole surface,
and in alcoholic specimens appear to be united together in
horizontal series by thin folds of the wall (fig. 14 0).
Panama; Paita, Peru; San Salvador.

Explanation of figures.

Fig. 7. Haloclava producia (St.). Type, natural size.

Fig. 8. Anthoplewra Dowii V. Side view of atentacle, actinobranch, and two
upper verruce. Enlarged.

Fig. 9. A similar view of the same organs of Anthopleura Stimpsoni V., from
the type, enlarged.

Fig. 10. Asteractis Bradleyi V. Type, about vatural size.

Fig. 11. The same, side view of tentacles and actinobranchs.

Fig. 12. The same, view of a portion of the disk, tentacles, pseudofronds,
enlarged.

Fig. 13. Eucladactis grandis Ver. Type, $ natural size.

Fig. 14. The same; a, one tentacle and a single vertical row of actinobranchs ;
b, a group of actinobranchs or papille from the middle of the column; c, various
forms of marginal actinobranchs.

Fig. 15. Lebrunia Dane, one of the actinobranchs with acrorhagi, 4+ natural
size.

Figs. 8 to 15 are by A. H. Verrill. Fig. 7 is by Wm. Stimpson.

Erratum—In part I, p. 494, line 17, for W. E. Coe, read W. RB. Coe.

Hillebrand—Analyses of Tysonite, ete. 51

Art. VII.— Mineralogical Notes: Analyses of Tysonite,
Bastnisite, Prosopite, Jeffersonite, Covellite, etc.; by W.
F, HILLEBRAND.

In the following pages are given the results of several
analyses of minerals made in the laboratory of the U. 8. Geo-
logical Survey during the past few years. In the cases of
tysonite and bastnadsite the results show that the formulas
attributed to them are correct, which in the absence of fluorine
determinations could not hitherto be affirmed. In the other
eases the analyses are interesting only as affording additional
data regarding rare minerals from new localities.

Tysonite and Bastndsite.

These minerals formed a single fine specimen half as large as
the fist, without crystal faces, from Cheyenne Mountain, near
Pike’s Peak, Colorado. The bastnasite covered one side of the
tysonite to the depth of an inch. The line of demarcation
between the two minerals was sharp, but examination of their
sections by Mr. H. W. Turner showed the tysonite to be per-
meated by stringers of bastnasite along numerous cracks and
that occasional grains of the latter were imbedded in the
tysonite, which accounts for the CO, shown in the tysonite
analysis. Attached to the tysonite at portions of its surface
were other white and brownish alteration products derived
from it, as shown by qualitative tests. The tysonite was evi-
dently the remnant of a single large crystal, since, according
* to Mr. Turner, all parts had the same optical orientation. Mr.
Turner further found the optical properties of both minerals,
so far as determinable, to agree with those given in Dana’s
Mineralogy, and the index of refraction of the bastnisite to be
greater than that of the tysonite. He likewise noted in both
minerals minute colored inclusions, indeterminable and very
trifling in amount, and also in the tysonite “numerous minute
angular cavities in which there is a liquid, often with gas
bubble. Minute, clear, cuboidal crystals, apparently isometric,
were also noted in some of these cavities.”

The composition of the minerals was found to be as given by
Allen and Comstock, with the exception that the ratios
of cerium oxide to the oxides of the lanthanum group are not
quite the same. The formulas are not thereby affected.

Cerium was separated from the lanthanum group oxides by
two precipitations by potassium hydroxide followed by long
introduction of chlorine. After recovery of the earths remain-
ing in solution, they were again subjected to this treatment to
be certain of having all the cerium. In one case a small por-
tion was thus recovered. The cerium was most carefully
examined for thorium and traces of what appeared to be thoria

52 Hillebrand—Analyses of Tysonite, Bastnasite,

were found. The other earths were wholly precipitable by
potassium sulphate with exception of traces of what may be
oxides of the yttrium group. Approximate molecular weight
determinations of the combined oxides of these two groups
were made, and they show an appreciable difference, which
may, however, be due to the uncertainty of the method. — It
may be mentioned that on ignition of the sulphates of these
earths they acted like the old didymium in that they lost
exactly two-thirds of their SO, on ignition over the full flame
of the Bunsen burner, a fact which would seem to exclude the
presence of lanthanum. Their solutions were pink and gave
pronounced absorption spectra.
cerium were a dull dirty brown, which became nearly white
on blasting and acquired a distinct bluish cast on ignition in
hydrogen. No appreciable reduction in weight followed heat-
ingin hydrogen. The material saved is at the disposal of any
one desiring to examine these earths spectroscopically.

Owing to the great difficulty in effecting complete decompo-
sition of the minerals by sulphuric acid at a single treatment,
the fluorine was obtained in condition for estimation by fusing
with potassium carbonate after mixing with silica in the pro-
portion of 06 gram mineral to 1 gram silica. 3

Fragments of tysonite when held in the blast gave a distinct
crimson flame showing the lithium red line, but an alkali
determination failed to reveal more than a trace of this element.

Of the bastnasite very little pure material could be sepa-
rated, and it was therefore impossible to place with certainty all
the loss shown by the analysis, but a portion of it is to be charged
to the oxides of the lanthanum group because of an accident.

Sp. grav. of the bastnasite 5-12 at 27° C. and of the tysonite
6:10 at 28° C., which becomes 6°14 when corrected for 2°65
per cent of bastnisite.

Analysis of Tysonite. Analysis of Bastnasite.

The ignited oxides freed from.

COCh een 42-89% GeO. (ius Gee aT 71*
La group.---- 39°31 La group -..-.-. 36°29+ (low)
Baie ehyien 28-71t Boe. io ee
C0), aia 53 TC 20:03t
02) 0) eee een "18 ie, O. . Aes [29
Hen be 22. 23 ald Nia, O§ . soe 18
Wa OS See 42), *SO0:(approxa). wie ©.) see 08
112°08 102°34
Less O for F, 12:08 Less O for F... 3°30
99°95 99°04

* 13 per cent ThO.?
+ At. w. 139°7; includes
°21 per cent soluble in K2SOx,.
Mean of 28°86 and 28°56.
With traces of K and Li.

* Mean of 37°73 and 37°69; in-
cludes ‘10 per cent ThO,?

+ At. w. 141; includes °09 per
cent soluble in K,SQx,.

t+ Mean of 19°94 and 20°11.

§ With traces of K and Li.

Prosopite, Jeffersonite, Covellite, ete. 53

Neglecting the three last constituents in each case, the ratios
become for

Bastnasite.--. R: F, CO, = 1: 2°94
Tysonite..-.. R: EF, CO. ==) W805

which ratio for tysonite is not changed by allowing for admixed
bastnasite.

The above direct fluorine determinations fully establish the
hitherto assumed formulas, RF,’ for tysonite and R’”’’(F,’CO,”)
for bastnasite.

Prosopite.

Over two years ago Mr. Geo. F. Kunz sent for examination
a beautiful pale green mineral from Utah, supposed to be iden-
tical with the green variscite called by him utahlite in Mineral
Resources of the U.S., 1894, p. 602. Under a recent date
Mr. Kunz writes that Mr. T. H. Beck of Provo found the
mineral “in 1895, in the Dugway mining district, Torvel Co.
It was found in a low range of hills about five miles long, sur-
rounded by a desert on an arid region occurring as flat rock,
associated with fluorite, native silver and slate, and trachytic
rock (?), containing decomposed. pyrite in which there was
present a little free gold.”

Unexpectedly this was found to be the hydrous aluminum-
calcium fluoride prosopite, mixed with some quartz and proba-
bly fluorite and colored by a small amount of some copper salt.
‘A new and interesting occurrence for this very rare mineral is
thus afforded.

The material as prepared for analysis after separation by a
heavy solution proved to be still far from pure; quartz grains
in amount from one to two per cent were left undissolved after
complete conversion of the fluorides into sulphates and pre-
sumably considerably more had been removed by the escaping
fluorine. The total amount of quartz was not determined and
the material at hand did not suffice for attempts at more com-
plete purification, so that the conclusions drawn from the
analysis, while extremely probable, are not to be taken as alto-
gether proven.

The sp. gr. of the mineral as analyzed was 2°87 at 21° C. and
the hardness about 4°5, both agreeing with the constants for
prosopite. Furthermore, but little of the water (1°25 per cent)
was expelled by several. hours heating at 280°C. Analysis
gave ;

54 Hillebrand—Analyses of Tysonite, Bastnasite,

WAN! eLGE OR e ee S 20°08
Canty. ee iia ais eee 17°55
Mo sac ycee, eee trace
Kove cu BARE a Se 12
IN a! ka els Ae ara "32
5 Reperoe yee kromrren VIRB EeN 1h
Bg 2G, ee ei) pata 28°00
THO. 32 eae es 14°24

Quartz and oxygen -.- 19°5

100°00

Neglecting copper, alkalies, and the oxygen calculated for
their oxides, and assuming the water to exist entirely as
hydroxyl, the following not very satisfactory atomic ratios
result ;

Pilipe aide. <i 28 Gs aaa A "7407 2
Caleta, : Sophie "4380 1°18
Ee eee 2+ asian 1:4690
Hydroxyl ......- bani sa:
which become
GANT aReS a SEE SENS F407 2
Cavin BOre rin 3712 1
ele feiss sehle, 13354
Hydroxyl --.---- 1°5808 oe

if enough calcium and its equivalent in fluorine are subtracted
to make the ratio Al: Ca exactly 2:1, on the not improbable
assumption that fluorite is present as an admixture, an assump-
tion that had to be made also for the Colorado prosopite in
order to bring it into close agreement with Brandl’s formula.

There is now a deficiency in the acidic radicals. The
figures for Al, Ca, and H,O are undoubtedly very nearly cor-
rect while the fluorine may well be half a per cent low, having
been determined by the Berzelian method, owing to the diffi-
culty of securing complete decomposition of the fine powder
by a single treatment with sulphuric acid. Let it be permitted
to balance the basic and acidic radicals by raising the fluorine,
and to figure the ideal percentages on this basis. These
become of interest when compared with the corresponding
figures for prosopite from Altenberg and Pike’s Peak as given
below: ,

Altenberg. Pike’s Peak. . Utah,
2's be 15 od 23°37 22°02 22°74
Cart i t aan 16°19 17°28 16°85
A Suey ig Eee 35°01 33°18 29°95
EO sees 12°41 13°46 16°12
Oe 2 a8 13°41 14°34

» 100:00

Prosopite, Jeffersonite, Covellite, ete. 55

If the assumptions made in the foregoing are justified, the
Utah mineral is prosopite, and further evidence is afforded of
the correctness of the view established by Penfield that fluorine
and hydroxyl can mutually replace one another in many min-
eral species, for their relative proportions differ materially in
the prosopite from the three known localities. The correctness
of the formula as applied to the Colorado and Utah prosopite
is, however, predicated, as said, on the unproven assumption
that the material analyzed contains some admixed fluorite.

Jeffersonite.

Two brown substances associated with franklinite and other
zine minerals from Franklin Furnace, N.J., so alike in appear-
ance as to have been taken for the same mineral species, were
received from Mr. Geo. L. English. One wasa little duller than
the other and proved to bea mixture of several minerals accord-
ing to Mr. F. L. Ransome of the Geological Survey, largely
pseudomorphic after some micaceous mineral. From Prof.
Clarke’s calculations, based on the following analysis, it might
be a mixture of a calcium-aluminum garnet, troostite, and
limonite. SiO, 32°09, Al,O, 11°12, Fe,O, 5-16, MnO 15°85,
ZnO 16°89, CaO 15°65, H,O 2°15, MgO and alk 1:12; total
100-00. |

The other was of a richer and deeper brown and showed
such a pronounced cleavage or parting in one direction as to
produce a lamellar structure. The luster was brilliant on these
cleavage surfaces. Other directions of cleavage were apparent.
The hardness was about 5:5 and the density 3°39 at 21°5° C.
Before the blowpipe a fragment fused with difficulty to a light
colored blebby glass. Analysis gave:

i ake. eee 51°70
eee.) eas 36
He Os 2. See eae
ICY GP eae pees Ge 7°43
PIS... 3°31
CA 32... 2 eee 23°68
MeOQrys... . 9g Iss 12°57
RO 2... ee 1b
PO ee) <a trace
ie. 22... eee OS

100°19

TiO,, FeO, P,O, absent.

Neglecting the sesquioxides, alkalies and water, this leads to
the ratio Si0,: RO = 1:1:02, and the formula is that of a
metasilicate R’’SiO,. According to Mr. English the material

56 Hillebrand— Analyses of Tysonite, Bastnasite,

submitted by him had been pronounced by Prof: Penfield on
the basis of qualitative tests to be jeffersonite, a manganese-
zine pyroxene, a statement supported by the analysis above
given, although neither the color of the mineral nor its quanti-
tative composition agree with the hitherto published data. In
Dana’s Mineralogy the color is given as “ greenish black, on
the exposed surface chocolate brown,” the density as 3°36 on ~
p- 358, but 3°63 on p. 360. The discoverers of the species,
Keating and Vanuxem, give 3°50-3°55 for the density and 4:5.
for the hardness. The present mineral presents all the evi-
dences of being fresh and unaltered, yet it is brown through-
out, and its analysis furnishes figures widely at variance with
those of Herrmann and of Pisani, but giving a better meta-
silicate ratio than either of their analyses. Notwithstanding
these discrepancies, there is no reason for ascribing to the min-
eral a new subspecies name. The analysis is chiefly valuable
as showing a wide range of composition for the mineral.

Covellite, Hnargite, Stalactite.

The only important occurrence of covellite in this country is
at Butte, Montana, where it occurs in splendid indigo-blue
masses. Specimens from the East Greyrock mine, collected
by Mr. Geo. W. Tower of the Geological Survey, gave almost:
the theoretical composition as shown below. Sp. gr. at 26° C.
4-76, uncorrected for impurities.

An analysis of enargite collected by Mr. Tower in the Rarus. '
mine, and of a beautiful sky-blue stalactite from the Anaconda
mine, both at Butte, are likewise given.

Covellite. - Enargite. Stalactite.
Cu -- 66°06 Ga 22° 48°67 CuO _.. 9732
Dees 33°8y He--22.--- #33 FeO _. 18.
He-es. | 14="30Mes, Ap a 210 MeO. 27-08
Insol. ALI AS hae 79d ALO, -.) 40°67
Sb. 2. 176 SO, -.. 35°05
100°18 $< 31°44 PO. Sas
Ratio Cu:S as 1:1:01 Insoles dal AsO) econ,
H,O... 43°44
100°32 Insol. - . 06
100°00:

The stalactite was of some size, and was most readily soluble
in cold water, the solution giving a strong acid reaction. The
calculated ratio is strongly acid, showing either a mixture of
highly acid salts or of normal salts with free acids.

Prosopite, Jeffersonite, Covellite, etc. 57

Fibrous sulphate.

Mr. W. H. Weed collected in the St. Paul mine, near White-
hall, Montana, a magnificent specimen of a compact soluble
fibrous sulphate, supposed to be melanterite. It seemed to be
a filling between fragments of broken rock. Outwardly it
was white from dehydration, but at some depth the unaltered
green mineral was to be found. This had the following com-
position :

AO de OO) G8) apa ee hs 4:34
BO) 2 161 0) 6) | 98 a IO Ged
10 deel tees se aremag ies Rel Raat 03
MNOS 20055 ei, Sed 2°62
Pn@, Os els 7) ser eee ee 1°06
COON. sah dca ts ba Bae Bae 05
Cars bite eS & ee ce ‘09
Mir Oe $5... 2 Sek eee sot 3°07
Ne Oe le oH Tig Ae OTS
> 0 Met A RS as 2 ae 29°88
UE COP a) 52 nay eee 2 none
Eos ORE PRS Cap ner hoe atl 48:84
arene gs a ee ae ee eS) EG
99°25

Of the water 10 per cent escaped in 24 hours over sulphuric
acid and only ‘4 per cent more in another like period, but a
total of 14:4 per cent after 10 days uninterrupted exposure.
The water thus lost is very slightly reabsorbed on exposure to
air.

From calculations by Prof. F. W. Clarke the substance may
be regarded as a mixture of alunogen or the halotrichite group
with salts of the melanterite group, the cmp ieee formula
being nearly

(Fe, Mn) ,(Zn, Mg), Al,(SO,), . 65H,0.

The outer white zone of the specimen contained only 39°62
per cent of water.

Laboratory of the U. 8. Geological Survey, July.

58 EF. W. Sardeson— What is the Loess?

Art. VIII.— What is the Loess? by F. W. Sarpuson.

THERE is wide difference of opinion regarding the origin
of the American loess. Is it an aqueous or an zolian deposit ?
Many believe it to be partly aqueous, partly solian, and thereby
raise the very important question of how to distinguish the one
kind of loess from the other. To those who view the loess as
geolian, this discrimination is easy, for the loess is all eolian
except locally where distinct sedimentary characters appear.
The loess is interpreted by them as wind-carried dust, deposited
on a land surface interrupted by streams and ponds.

The loess viewed as a sediment is also easily distinguishable
from other sediments, but requires then the very difficult
explanation of how a little altered glacial sediment could be so
peculiar,—in fact like dust deposit. Negative evidence to this
theory is found in a region like Minnesota, where there are
thousands of existing and extinct lakes of glacial origin, in not
one of which is loess known to have deposited,—not even in
Lake Agassiz. They have in them gravels, sand and silt, and
around them are beaches. No small nor shallow lake could
have spread over the extremes of altitude on which the loess
lies, and a necessarily huge body of water must have left some
beaches and coarse sediments besides silt or loess. It was not
a lake that deposited the loess. The loess could also scarcely
be all a glacial sheet flood deposit. or example, the Iowan
loess lies beyond the Iowan drift sheet as it were its continua-
tion. Either the loess is the older and hence remains intact
only beyond the Iowan glacier’s domain, or else the loess is con-
temporaneous and deposited by agency of the same glacier’s
waters. The latter is the sedimentary theory. But the theory
does not explain why the glacial waters ceased depositing
loess when the glacial retreat began. Lowan loess does not
cover the lowan drift,* and this fact the szolian theory alone
can explain. Partial or entire recognition of the eolian origin
is necessary.

Partial recognition does not, however, remove all incon-
sistencies, because “no means of discriminating between the
two kinds of loess are yet known to be formulated,” as the
advocates of the mixed hypothesist say. Prof. T. C. Cham-
berlin teaches this double origin of the loess, and his view, as
understood by the writer, is that the glacial waters flooded

* Prof. 8. Calvin discredits the reported slight occurrence of this loess on the
Jowan drift. Iowa Geol. Sur., vol. viii, p. 174; see also p. 339.

+ C. R. Keyes, this Journal, vol. vi, p. 229.

} T. C. Chamberlin, Jour. Geol., vol. v, p. 795.

EF. W. Sardeson— What is the Loess ? 59

sediments upon the land, and these were in part modified by
zeolian agent into “ bluff loess.”

in theory, the Chamberlin view may be essentially correct.
The loess particles came from the drift, and very probably
from washed drift largely. If the theory be applied, it should
bring into recognition the following kinds and relations of
deposits: the unassorted glacial depssit, or till; this, when
assorted by agency of water, becomes stratified gravel, sand
and silt; the two latter. may be made by eolian agent into
dune sand and “bluff” loess; and they, after erosion, may
become again respectively stratified sand and silt ; under the
action of humic acids the silt of both kinds may be reduced to
a loam in which there is no lime; and likewise the “ bluff”
loess may be reduced entirely, or to a degree, to “loess” loam.
The till is easily recognized, as are also the stratified products,
the gravels and sands. Dune sand lacks the assorted and strati-
fied structure, and is easily distinguishable from the sedi-
mentary patches associated with or in it. It seems strange,
therefore, that silt and eolian deposit,—that is mud and dust,—
should not be distinguishable.

Indeed, it is not difficult to discriminate between glacial silt
and loess, and I believe that they are rarely if ever confused by
geologists. The difficulty lies in the discrimination of theo-
retical sedimentary loess and eolian loess. There are two
kinds of loess in the Mississippi basin, and they are distinguish-
able at a glance, but as said, they have not been proved separ-
able on the basis of sedimentary and eolian origin. The two
kinds are very intimately associated. They are respectively
the characteristic calcareous, porous, “bluff” loess and the
oxidized portion of the same, which may be called loess loam,
residuary loess, or ferretto. Difference in color, or the appli-
cation of hydrochloric acid, serves as a ready test of them.*
Chamberlin and Salisbury, in their monograph, recognized the
loess loam of the “ Driftless Area” as residuary clay although
they failed to connect it with the loess, as I have once before
indicated,t and in fact Chamberlin has practically admitted
the emendation by criticising other parts only of my paper.t
The loess loam being due to oxidation by humic acids, its
depth is determined by the vegetation upon its surface and by
the time that this agent has acted, provided that the loess be
not too shallow and surface erosion does not equal or exceed
the rate of oxidation.: On the Missouri and Mississippi River
bluffs, the loess was originally deepest; it is also little covered
with vegetation and highly exposed to erosion. On the higher

*This distinction is not new, for example see Geologischer Fuchrer d. Umge-

bung v. Freiburg, p. 78, Steinmann u. Graeff, 1890.
+ Am. Geol., vol. xx, No. 6. t Jour. Geol., vol. v, No. 6.

60 EF. W. Sardeson— What is the Loess ?

but more level land farther from the great rivers, the loess was
thinner and was also more oxidized. Often it is all oxidized.

From my own observations compared with published deserip-
tions, I feel safe in saying that, asa rule, the loess loam is what
has been called sedimentary loess. If it were such, why is it
not calcareous as are the known glacial silts? And if the
“ bluff ” loess is derived from it, why is the former calcareous,
since the latter is not? There is no explanation. 3

The above criticism applies to the- loess as a rock type, but
when the Loess formation is considered another distinction
should be admitted, namely that there are two distinct loess
formations, as H. F. Bain describes in northwestern Iowa,*
which are comparable to the older and younger loess of the
Rhine Valley. Each may bear loam.

The loess as a formation bears loam both upon and within.
Upon the loess, the loam represents the portion which has been
oxidized since the loess formation was deposited, while in the
same it indicates patches of vegetation that were contempo-
raneous with the deposition. The loess loam, in combination
with land and fresh water shells, forms the very strong argu-
ment in favor of the purely eolian origin of the loess. The
colian hypothesis is the one which permits the several parts of
the loess to be both readily interpreted and easily recognized
in the field, and it has therefore a distinct advantage over the
semi-eolian hypothesis. It is further competent to explain
without inconsistency the dune sands on the river bluffs and
their graduation into the loess; wind polished pebbles lying
between till and loess are recognized; two kinds of stratified
silts existing with the loess or graduating into it are identified
as respectively glacial silt from which the loess seems to have
been largely derived, and modified or washed loess. Banded
structure in the loess,+ which is not true stratification, indicates
the line of a moist zone where vegetation thrived and oxidized
the loess. Dry runs and very probably rodent burrows and
mounds can be easily recognized ; and the occurrence of land
and fresh water shells is not anomalous. Finally, when it is
observed that much loess accumulated before one glacial period
or ice advance and little before another, the same is consist-
ently attributed to climatic conditions, to which kind of cause
the glaciers themselves are referred.

University of Minnesota, Minneapolis.

* Towa Geol. Sur., vol. viii, p. 240.
+ See for example, fig, 32, p. 235, vol. vii, Iowa Geol. Sur.

Hutchins—Absorption of Gases in a High Vacuum. 61

Art. 1X.—Absorption of Gases in a High Vacuum; by
C. C. HutTcHins.

THE term “absorption” is here used to cover the general
phenomenon of the disappearance of the residual gas in a high
vacuum under the electric discharge, whatever may be the
cause of the disappearance.

Jt is well known to all users of X-ray apparatus that under
the discharge from an induction coil the tubes rapidly rise in
vacuum, and the current, after a longer or shorter time, ceases
to pass, and the usefulness of the tube is at an end. ‘This rise
in vacuum constitutes the most serious defect in such tubes
and is a source of constant annoyance.

It is probably safe to reject the theory that the electromo-
tive force drives the molecules of the residual gas through the
walls of the glass. It is possible to have an opening into a
highly exhausted tube without the entrance of air.

This laboratory has for twenty years possessed a Crookes
radiometer that has during all that time been cracked entirely
around near the top of the bulb, but which, notwithstanding, con-
tinues to function perfectly. We also had for several months
an X-ray tube similarly cracked. By constant use this tube
became so high in vacuum that it was punctured by a spark; air
entered and the cracked end fell away of its own weight.

The disappearance of the gas is therefore probably due to
mechanical and chemical combination with the metallic elec-
trodes of the tube, and it should be possible to find some gas
or vapor whose rate of absorption would be less in rate or
extent than that which takes place in the ordinary tubes.

The first step in the inquiry is to learn what the ordinary
high-vacuum tube contains.

If a Plicker tube, made from carefully cleaned glass, be
very highly exhausted and then the current from an induction
coil turned into it, the vacuum falls at once, air and water
being driven from the walls of the tube. If now by heating
the tube, continuing the current and working the pump, the
vacuum be again raised to the point at which the fluorescence
of the glass appears, the spectrum of the tube will be found
to consist of the red and blue lines of hydrogen (C and F,
and three or four lines of oxygen; that is, the tube contains
only water. The tenacity with which glass retains water upon
its surface is well known ; and it is necessary to heat the glass
nearly to its softening point before the last of the water comes
away. When the vacuum in a tube is so high that the current
will not pass, heating the tube will dislodge a little water and
the current will pass until the water has again condensed upon
the cooling of the tube.

62 Huitchins—Absorption of Gases in a High Vacuum.

As the following experiments had for their object the pro-
duction of a permanent state of vacuum under the electric dis-
charge, they were limited to such gases as could be produced
readily in the tubes themselves, and whose use seemed to hold
ont promise of success.

A number of Plicker tubes were made, as nearly alike as
possible, and each provided with a small communicating tube
containing some chemical that would yield the desired gas by
the application of heat. The chemicals were prepared as pure
as possible, and dried at a hundred degrees for twenty-four
hours. The dried-out tubes, containing the chemicals, were
then attached to the mercury pump and exhausted; heated
very hot and pumped to remove all water vapor. The tube
containing the chemical was then heated, gas driven off into.
the tube and the exhaustion repeated. |

In nearly every case it was found impossible to entirely get
rid of water vapor, even by repeatedly washing out with the
gas. The tubes were sealed at such a degree of exhaustion as
would allow the subsequent rise in vacuum to be well marked,—
when the green fluorescence of the glass appeared at the
cathode, and the sparking distance across the terminals of the
coil and in parallel with the tube was 2™™. After the tube was
sealed, it was placed before a spectroscope, and the current
from the coil passed until the spectroscope showed the entire
absorption of the gas, or until the vacuum became too high
for the current to pass readily. A fairly good idea of the rela-
tive rates of absorption of the various gases may be obtained
from the time required for the absorption to take place.

Tube containing zinc cyanide.— This substance yields
cyanogen when heated. The tube was washed out with
the gas as above described and sealed. The vacuum began
to rise at once as soon as the current was turned on. In
five minutes the whole tube became fluorescent, and in
seven minutes the cyanogen was entirely absorbed, the spec-
trum being now that of the trace of water vapor from which
the zine cyanide could not be freed. . It is difficult to see why a
gas which if decomposed by the discharge would remain as the
chemically inert nitrogen, should disappear at such an extra-
ordinary rate. To test the matter further another tube was
made in such form that when the vacuum became higher than
a certain value, a portion of the current was shunted through
a platinum wire surrounded by zine cyanide, and so the vacuum
was automatically maintained nearly constant. This tube was
worked for four hours, and the yield and absorption of the gas
went on for the whole time at a nearly constant rate.

Tube containing lead ferrocyanide, yielding nitrogen when
heated, also gave considerable water vapor. Sealed at
a slightly lower vacuum than the first, yet the nitrogen all

Hutchins—Absorption of Gases in a High Vacwum. 63

vanished in ten minutes, water alone remaining. As it
was possible in these two cases that the gases had under-
gone reabsorption upon the cooling of the substances from
which they had been expelled, a fresh portion of the gases
was driven off and the tubes allowed to stand twenty-hours.
No change could be observed in them.

Tube containing codine.—The iodine was shut off from the
main tube by a stopcock, fearing its vapor tension at ordinary
temperatures would be too great to admit of obtaining a suffi-
ciently high vacuum. The tube was exhausted, the stopcock
opened for a moment to fill the tube with iodine vapor, and
the exhaustion repeated. After sealing, the iodine disappeared
under the action of the current in less than a minute. The
tube was filled with iodine vapor again by opening the stop-
cock and warming the tube. Upon again starting the current
the platinum mirror upon the glass at the cathode was seen to
be attacked and converted into a dark bronze-colored substance
and the iodine again disappeared.

Tube containing o«ide of mercury, yielding oxygen when
heated, but no water vapor. Tube pumped and heated very
thoroughly, and when sealed showed the spectrum of oxygen
and the red hydrogen line faintly. The vacuum rose slowly ;
after fifteen minutes the hydrogen line disappeared ; at the
end of an hour the vacuum had become quite high, the spark
in parallel with the tube being two inches, but the current still
passing easily. The spectrum of the tube then consisted of
three lines of oxygen.

Experiments were also made with carbon dioxide and some
other gases, the details of which need not be given here, inas-
much as it was found that oxygen was absorbed far more
slowly than any other substance tried. Mercury oxide is also
eminently adapted to maintaining a condition of constant
vacuum. It yields oxygen only at a high temperature, and the
flow of gas ceases immediately upon withdrawal of the source
of heat. |

To construct an X-ray tube in which the vacuum may be
maintained at the most efficient point, it is only necessary to
attach near the anode terminal a short tube having a sealed-in
platinum wire packed around with oxide of mercury. A plug
of asbestos may be used to hold the substance in place. When
by use the vacuum becomes too high, a portion of the current
is shunted from the cathode through the oxide of mercury for
a brief time; a little gas is dislodged and the tube becomes as
new. The shunt may be left permanently in place with a
longer or shorter spark-gap at the mercury terminal. The
action will then be automatic, and vacuum remain indefinitely
at any desired point.

Searles Physical Laboratory,
Bowdoin College, November, 1898,

64 Scientific Intelligence.

SCIENTIFIC INTELLIGENCE.

I. .CHEMISTRY AND PHYSICS.

1. On the Extraction of Nickel by the Mond Process.—A paper
by Rozsrerts-AustTEN has recently appeared giving an account of
the application of the volatile nickel-carbonyl, discovered by
Mond, to the extraction of nickel from its ores. Mond’s discovery
was that metallic nickel forms a volatile compound with carbon
monoxide, which he called nickel-carbonyl, which boils at 150°
and from which the nickel may again be regenerated by heating
to 180°. Iron, while not as active, acts similarly. The sugges-
tion was an obvious one, that by means of these volatile com-
pounds, these metals might be separated from eobalt, copper and
the others with which they are associated in their ores. An
experimental plant was erected in 1892 near Birmingham. The
material used was Bessemerized matte, which after roasting con-
tained 35 per cent nickel, 42 per cent copper and about 2 per cent
of iron. By treatment with sulphuric acid, about two-fifths of
the copper was removed, the residue containing 51 per cent of
nickel. The remaining copper was reduced to the metallic state
by means of water gas at 300°. The ore was then treated in a
volatilizing tower with carbon monoxide, the temperature being
kept below 100°. The volatile nickel carbonyl passed to a
decomposer—a horizontal retort heated to 180°—where the nickel
was released in the metallic form and the carbon monoxide
returned to the volatilizer to be again used. The unused metallic
residue was returned to the roasting furnace. The nickel produced
was 99°8 per cent pure. Up to the time Professor Roberts-Austen
visited the plant, about 80 tons of nickel had been thus prepared,
with quite satisfactory results.— ature, lix, 63, November, 1898.

Gi ae

2. On Atherion.*—A note on extherion has been published by
CROOKES, giving some of his old researches made from 1873 to
1881, tending to show that the new gas may be nothing more
than aqueous vapor. Early in his investigations he noticed that
a little aqueous vapor thrown into the vacuum [of a radiometer],
converted into attraction the repulsion due to radiation. In 1880,
he observed that aqueous vapor was found to retard the force of
repulsion to a great extent and carbonic acid acted in a similar
way, though less in degree. ‘The presence of even a trace of
aqueous vapor was found to have a strong action in diminishing
the sensitiveness of the radiometer and other instruments,” he
says. As to the absorption of etherion by glass and its evolution
again on heating, the author gives an experiment made in 1879
showing that air-dry glass condenses on its surface a considerable
quantity of water and carbonic acid which does not come off

* This Journal, IV, vi, 431, November, 1898.

Chemistry und Physics. 65

until heat is applied. As to its conductivity for heat, experiments
are described which show that at high vacua water gas is a better
conductor than either air or hydrogen at similar pressures. “It
has been found,” he says, “that as water gas is rarefied, its heat
conductivity diminishes in a greater ratio than that of hydrogen.”
«At these high vacua water gas gains so rapidly on hydrogen
that I am prepared to believe that at Mr. Brush’s low pressure of
0°38 millionth, the ratio may even be as great as that he ascribes
to etherion.” ‘‘On the evidence at present available,” he con-
cludes, ‘‘ I consider it more probable that ztherion is water vapor
than that it is a new elementary gas.”— Chemical News, \xxvili,
221, November, 1898. Cukr Bi

3. On the Preparation of Graphitic Acid.—An easy method
of preparing graphitic acid has been described by STauDENMAIER.
A mixture is first made of 100° of concentrated nitric acid (sp.
gr. 1:4) and 300° of ordinary strong sulphuric acid. ‘To this are
added 50 grams of pulverized Ceylon graphite, and then 100
grams of potassium chlorate, in small portions at a time ; the
whole being frequently stirred. After the mixture has stood for
several hours, it is poured into water, and the solid is washed,
dried and ignited in a large dish of metal. until it has intumesced.
Then it is again stirred with water, and the portion that floats is
collected for use in the subsequent operations. Of this prepared
graphite, 25 grams are poured, with constant stirring, into a cold
mixture of one liter: of strong sulphuric acid and half a liter of
strong nitric acid contained in a shallow dish, and then about 450
grams of potassium chlorate are added in successive portions.
After the active evolution of gas diminishes, and a sample of the
resulting green substance appears pure yellow on treating it
with acid permanganate, the whole is poured into water, and the
resulting solid is washed repeatedly by decantation. It is now
placed in a dish and a solution is added to it consisting of seven
grams of potassium permanganate dissolved in 120° of hot water ;
after cooling, a mixture of 15°° of strong sulphuric acid with 75°
of water is added, and the whole is heated on the water bath
until the red color has disappeared. Hydrogen peroxide is
added, the whole is allowed to stand at rest for some time with
occasional stirring, and the graphitic acid is then washed first
with dilute nitric acid and finally with alcohol and ether.—er.
Berl. Chem. Gres., xxxi, 1481-7, June, 1898. Gi ELE

4. On Compounds of Lithium and Calcium with Ammonium.
—The fact has been observed by Motssan that when liquefied
ammonia, contained in a tube cooled to —50°, is brought in con-
tact with metallic calcium or lithium, solution takes place with
the production of an intense blue liquid having a brown red
reflection ; this result being comparable to the similar phenome-
non observed with potassium and sodium. These latter metallic
ammoniums when thus produced, according to Joannis, have the
formulas KNH, and NaNH,, and they decompose at the ordinary
temperature and pressure, evolving ammonia gas and regenerating

Am. Jour. Sci.—-Fourta Srrizs, Vou. VII, No. 37.—JAanuarRy, 1899.
5 .

66 Scientific Intelligence.

the alkali metal in brilliant crystals, as Seely long ago pointed
out. As opinions differ as to whether these metallic ammoniums
are molecular or atomic in constitution, the author has made
experiments with potassium, sodium, lithium and calcium. Each
metal was placed in a U-tube of glass, one leg of which was
drawn out fine, while the other was recurved and connected with a
source of ammonia gas, dried first over fused potassium hydrate
and then over sodium. The four tubes were united successively
so that the same current of ammonia gas traversed them all. B

the use of a water bath, the temperature could be regulated, and
the flow of gas increased or diminished by means of a tap. For
low temperatures acetone was used in the bath, being cooled even
as low as —75°, the solidifying point of ammonia, by adding to
it fragments of solid carbon dioxide. The experiments were
made under atmospheric pressure. It was found that lithium was
attacked at +70° with liquefaction, calcium at +20° without
liquefaction, potassium at —2° with liquefaction, and sodium at
20° with liquefaction. It is clear therefore that these four metals
- unite directly with ammonia gas, the temperatures obtained being
the superior limits of the action and at the same time the decom-
position-temperatures of the metallic ammoniums under the
conditions. By preparing the metallic ammonium by the action
of the liquefied gas, and then slowly raising the temperature, the
inverse method was effected, the sodammonium dissociating at
—20° and giving the metal and ammonia gas, and the potassam-
monium doing the same at —2°; while calcium-ammonium and
lithium-ammonium were stable at the ordinary temperature and
pressure. Moreover, ammonia gas solidified at —80°, does not
attack either of the four metals, though action takes place as soon
as liquefaction commences with evolution of heat. If the experi-
ment is made slowly, the author states that amides of these
metals are produced; while when it is rapid, the metallic ammo-
nium may be readily produced at the ordinary pressure and the
metal regenerated in crystals without the production of amides.
Lithium placed in a U-tube at the temperature of the laboratory
and subjected to a current of ammonia gas, takes a brownish-red
tint and liquefies, producing a liquid of the same color. Heated
rapidly to + 70°, the excess of ammonia escapes, leaving a reddish-
brown solid, which takes fire in contact with the air. If lithium
be placed in liquefied ammonia, the tube being allowed to return
slowly to the ordinary temperature, the dark blue liquid becomes
thicker, reaching a sensibly constant composition after 24 hours.
The composition then was found to be: NH, 89:04, 87:00, 88°72,
and 88°37 in four samples; the formula (NH,),Li requiring 87°93.
The liquid may therefore be considered either as a saturated solution
in ammonia gas or as acompound NH,Li.(NH,),. As to solid lith-
ium-ammonium, its composition was determined (1) by weighing
the ammonia fixed by the metal and (2) by preparing the com-
pound and determining the ammonia in it. The lithium found in
five experiments was 28°07, 28°40, 28°72 and 28°82 per cent.; the

Chemistry and Phystves. 67

ammonia being 71°93, 71°60, 71°28 and 71°18. Theory requires for
NH,Li, 29°16 of lithium and 70°83 of ammonia. It is much less
soluble in liquid ammonia than the sodium compound. The metal
calcium exposed at 15° to a current of ammonia gas takes a yel-
lowish-brown tint, becomes heated and increases in volume. Ata
temperature between 15° and 20° only a solid compound is formed,
which takes fire in contact with the air and which fixes liquid
ammonia and becomes pasty, being then soluble in the liquid.
Like lithium, it gradually decomposes at the ordinary tempera-
ture, yielding transparent crystals of amide (NH,),Ca and evolv-
ing hydrogen. On analysis calcium-ammonium gives numbers
agreeing with the formula (NH,),Ca.—C. &., cxxvii, 685-692,
November, ‘1898. | G. F. B.

5. On the Spectra of Lodine.—The results of a measurement
of the lines in the emission and absorption spectra of iodine from
wave-length 3030°5 to wave-length 6191, expressed in Angstrom
units, have been given by Konen. As is well known, either a
line spectrum or a band spectrum can be obtained in a vacuum
tube. The latter is composed of two different portions which
may be designated as B, and B,; the former of these, Bg, is iden-
tical with the absorption band spectrum, while By is not. When
the discharge is very intense, however, By disappears, while By
increases in intensity ; so that the author is inclined to the ae
that Bz is due to the ordinary iodine molecules, and By, results
from dissociation. In the line spectrum also, two series may be
distinguished, but it is quite possible that in this case the second
series may be produced by an impurity. Tables of the wave
lengths are given in the paper.—Ann. Phys. Chem., II, |xv, 257-
286, 1898. G. F. B.

6. A Manual of Chemical Analysis, Qualitative and Quanti-
tative; by G.S. Newtu, Demonstrator in the Royal College of
Science, London ; 8vo, pp. xii, 462. New York, 1898 (Longmans,
Green and Co.).—The author tells us in the preface to his book
that he has “done his best to make it as little of a cram-book as
possible, but has endeavored to teach analytical chemistry as well
as analysis—that is, the theoretical as well as the practical side of
the subject.” Book I treats of Qualitative, Book II of Quantita-
tive Analysis. The quantitative work is divided into gravimetric
methods, including electrolysis, and volumetric methods, includ-
ing gas analysis; special methods following these. The great
convenience of having the leading facts well set forth in a work
of moderate size, for the use of students taking short courses, is
well exemplified in this manual. The portion on volumetric work
seems especially good. G. FB.

7. Stratified Brush Discharges in Atmospheric Air.—M.
TorruLerR has described a species of stratification in brush dis-
charges (Wied. Ann., lxiii; p. 109, 1897) and the present paper is
a continuation of his studies in this direction, A sixty-plate
Toepler-Holtz machine was employed, and it was found that the
increase of potential-difference was closely proportional to the

68 Scientific Intelligence.

increase of sparking distance; and that the potential difference at
the electrode is nearly independent of the current strength if the
discharge is of the natureof abrush. The G. Wiedemann-Hittorf
law for Geissler tube discharges therefore holds also for brush
discharges in free air.— Wied. Ann., No. 12, 1898, 660-675.

The writer has substantiated the results of Toepler, as far as
relates to the stratification, by the employment of an apparatus
which can produce a difference of potential of three million volts.
The stratification with this high voltage occurs in the neighbor-
hood of the positive pole. Ja Tages

8. The Spectrum of Lighining.—Photographs of the light of a
spark six feet in length, taken in the Jefferson Physical Labora-
tory, show atmospheric lines very strongly developed between EH
and H. No more lines were produced by a voltage of three
million than can be brought out with a voltage of one hundred
thousand. With the higher voltage, however, there is a complete
absence of the metallic lines of the terminals. The photographs
taken with the high voltage therefore must closely resemble the
spectrum of lightning. 5 Pe?

9. Dispersion in the Hlectrical Spectrum.—Observations on di-
electric constants for periods between wave lengths 2° to 75°™ have
hitherto been wanting. KE. Marx has endeavored to supply this
break in the subject and has applied to his results certain laws of |
dispersion. He speedily ascertained that in order to obtain very
short electrical waves it is necessary to so interlock the primary
and secondary circuits, or in other words the exciting and reso-
- nating circuits, in such a manner that only the fundamental vibra-
tions of the primary circuit were in evidence. He employed
Drude’s method of measurement, which consists in nodes measure-
ing the displacement of the electrical waves on wires, when a
portion of these wires are surrounded by a dielectric. By means
of an exciter of small dimensions he obtained electrical waves,
4, 36°", and 58°™ in length. ‘The positions of the nodes were
ascertained by means of a Zehnder-Geissler tube. The index of
refraction of electric Waves, 3°2° in length, for water was found
to be

Tn = )850 (ati sc.)

For wave length 36°

n? = 82°50 (at 17° C.)
For wave length 53

1 = 188.73 watt CS)
The indices of refraction of ethyl-alcohol were similarly inves-
tigated. It was surmised that absorption bands of great extent
in the ultra-red modify existing dispersion formule. A difference
in concentration of the ethyl-alcohol solutions of 1 per cent can
change the square of the electrical index of refraction 35 per cent.
— Wied. Ann., No. 11, 1898, pp. 411-484; ibid., No. 12, 1898, pp.
597-622. Jeans

10. Traité Blémentaire de Méchanique Chimique fondée sur la
Thermodynamique par P. Dunrem. ‘Tome iil, Les Melanges

Geology and Mineralogy. 69

homogénes; les dissolutions, pp. 1-380. Paris, 1898 (Librairie
Scientifique A. Hermann).—The third volume of the extensive
work by M. Duhem, which was announced in volume iii, p. 419,
of this Journal, has now been issued. It shows that the promises
made at the outset are being thoroughly fulfilled, and that the
work when completed will occupy a unique place in scientific
literature.

The volume is divided into two parts; the first of these dis-
eusses homogeneous mixtures with respect to thermodynamic
potential, osmotic pressure and the hypotheses of Van t’Hoff and
d’Arrhenius; also chemical reactions in homogeneous systems.
The second is devoted to solutions, discussing the solution of salts
in general, the effect of vaporization of the solvent, with the
formulas of Kirchhoff, the effect of freezing of the solvent, saline
and gaseous hydrates, and double salts. This enumeration of the
subjects of leading chapters will serve to show the scope of this
part of M. Duhem’s work, but only close study will make the
reader familiar with the thoroughness of the discussion asa whole.

11. Prismatic and Diffraction Spectra. Memoirs of Joseph
von Fraunhofer. ‘Translated and edited by J. S. Ames.—The
inauguration of the series of Harper’s Scientific Memoirs was
announced in the last number of this Journal (p. 504, Dec., 1898).
The second volume has now been issued; it contains translations
of Fraunhofer’s classical papers (1817-1828). To these are added
the note by Wollaston (1802) describing the method by which he
obtained a pure solar spectram and discovered the presence of
fixed lines in it. This volume, like its predecessor, is edited by
Prof. Ames.

Il. GroLtocy AND MINERALOGY.

1. Maryland Geological Survey; Wi1u1am B. Cxrark, State
Geologist. Vol. i, pp. 1-539, plates i-xvii, 1897; Vol. ii, pp.
1-509, plates i-xlvili, figures 1-34, 1898.—Maryland was one
of the first four states in the Union to institute an official
geological survey, and, in the present series, is preparing the
most elaborate and elegant geological reports that any state has
so far produced. This second official survey of the state was
organized by act of the General Assembly in March, 1896. The
commission appointed Prof. Wm. B. Clark state geologist, and
the two volumes are the work of the first and second years of the
survey. The first volume, as the geologist says in the preface,
“consists primarily of a summary of past and present knowledge
concerning the physical features of Maryland, and embraces an
account of the geology, physiography and natural resources of
the state. The “ Historical sketch,” comprising Part ii, begins
with the landing of Captain John Smith, in 1608, and narrates
the work of the successive explorers and investigators who have
developed the knowledge of the resources of the state. Part iv,
on the Bibliography and Cartography, is compiled by Mr. E. B.

70 Scientific Intelligence.

Matthews and includes a list of the works and maps published
relating to the physiography, geology and mineral resources of
the state, from 1526 to 1896, inclusive, and Part v is the first
report by Mr. L. A. Baur on the magnetic work in Maryland,
including the history and object of magnetic surveys in general.

Volume ii, in addition to the state geologist’s administrative
report on the operations of the survey during 1896, and during
1897, contains in Part 11 an elaborate paper by Messrs. George P.
Merrill and Edward B. Matthews, on the Building and Decorative
Stones of Maryland, illustrated by beautitul colored pictures of
some of the more important stones as polished, and: micro-photo-
graphs of their structure. Part ili is an equally elaborate report
on Cartography, to which Mr. Gannett, of the United States
Geological Survey which is codperating with the state survey,
contributes a valuable paper on the aims and methods of carto-
graphy; and Mr. E. B. Matthews one on the maps and map-
makers of Maryland.

The volumes are printed on excellent paper, and the illustrations.
are numerous and often of great beauty as works of art, and noth-
ing seems to be spared to make the reports models of their kind.

H, S. W.

2. The Lower Cretaceous Grypheeas of the Texas Region ; by
R. T. Hitt and T. W. Vaueuan. Bull. No. 151, U. 8. Geol.
Survey; pp. 1-139, plates i-xxxv. Washington, D. C. 1898.—
At last the Gryphca pitchert Morton, and the age of the
Tucumcari beds, have been set at rest. Mr. R. T. Hill has
succeeded not only in solving, but in clearly demonstrating his
solution of the perplexing problems regarding the Lower Creta-
ceous beds of Texas and neighboring regions, in showing the
true order of the beds and their faunas. In this work, one of
the greatest difficulties has arisen from the great variability of
the Gryphzeas which have served as Jeitfossilien, and from the
confusion arising from mis-naming, and careless (or worse) state-
ments regarding their occurrence and distribution. Mr. Hill, with
the assistance of Mr. Vaughan, has made an exhaustive study of
the whole group of Gryphzas, and determined and illustrated
with numerous figures the characters of the species, their develop-
mental history and their stratigraphic range and geographical
distribution. The names of the Gryphzas (to which the name
G. pitchert has been applied) which survive the searching inves-
tigation are G. corruguta Say, G. navia Hall, G. mucronata
Gabb, G. washitaensis Hill and @. Newberryt Stanton. Two
new species, G. Wardi and G. Marcoui, are described. FPartic-
ular attention should be called to the ontogenic study of the
several species according to the Hyatt school of paleontologists.

H. S. W.

3. Bibliographic Index of North American Carboniferous
Invertebrates ; by Sruart Wetter. Bull. No. 153, U.S. Geol.
Survey, pp. ji 653, 1898.—The indexing and classifying of the
innumerable facts regarding fossil species is a necessary prep-

Geology and Mineralogy. 71

aration for the new paleontology, which deals with the history of
organisms rather than their form and description alone. Mr.
Weller has not only opened the way for his own investigation in
this direction, but has rendered a great service to all advanced
students who are engaged in paleontological investigation. The
matter is well arranged and the necessary statistics are given with
precision and will enable the student to turn at once to the original
sources of information. H. S. W.

4, Contributions to the Tertiary Fauna of Florida, etc.; by
Wma. H. Dati. Vol. iii, Part iv.—I, Prionodesmacea: Nucula to
Julia, II, Teleodesmacea: Teredo to Erville; (Trans. Wagner’s
Free Inst. of Sci. of Philadelphia), pp. 571-947, piates xxili-xxxv,
April, 1898.—This volume maintains the characteristics of its
predecessors ;—the beautiful, sharply-defined illustrations and the
clear and exhaustive descriptive parts, which are carried out in
many cases to a thorough revision of classification of the larger
divisions and the distribution of the known species, leave little to
be desired in the discussion of such a group of fossil mollusca.

H. Ss. W.

5. Contributions to Canadian Paleontology ; by J. FE. Wuir-
EAVES ; vol. 1, part v, pp. 361-436, plates xlviii-l; Geol. Survey
of Canada, No. 659, 1898.—This closing part of the first volume
of contributions to Canadian Paleontology is chiefly concerned
with the thorough revision of the nomenclature and faunas particu-
larly noted in the previous portions of the volume, the first of
which was issued in 1885.

In the first papers, the Hamilton fauna of Thedford, Widder
and neighborhood in Ontario is thus revised; the additions are
based upon new collections made by Mr. Schuchert, for the
National Museum, in 1895 and Canadian collections examined by
the author; and the revision of nomenclature is of particular
importance. as expressing the comparative study, of both Mr.
Whiteaves and Mr. Schuchert, on large and widely distributed
collections. The final list recognizes 226 species in the fauna.
The second number of the part is an appendix on revision of
nomenclature and statements in previous numbers of the volume.
The volume being now complete, a special title page is issued with
directions for binding the whole. H. Ss. W.

6. Geological Survey of Canada, G. M. Dawson, Director,
Ann. Rept. (new series), vol. ix, for 1896, pp. 816, five maps,
twenty plates, Ottawa, 1898.—In addition to the summary report
(614) of the operation of the survey for the current year 1896, the
volume contains Tyrrell’s report on the Doobaunt, Kazan and Fer-
guson rivers, etc. (part F), Bell’s report on the French river
sheet (I), Low’s report on the northern part of Labrador penin-
sula (L), Bailey’s report on S. W. Nova Scotia (M), Hoffmann’s
report on Chemistry and Mineralogy (S) and Ingall’s report on
Mineral Statistics. All of them except the Tyrrell report have
been previously noted (see this Journal, iii, p. 4215; iv, 78, 232,
434 and 510). H. S. W.

42 Scientific Intelligence.

7. Report on the Doobaunt, Kazan and Ferguson Rivers and
the Northwest Coast of Hudson Bay, etc.; by J. B. TyrrxE yu.
Geol. Surv. of Canada, Ann. Rept. 1896, vol. iv, part F, pp.
1-218, plates i-xi and three maps, 1897.—This report gives
account of explorations carried on during the years 1893 and 1894,
in a region lying north of the 59th parallel of latitude and west
of Hudson Bay, covering about 200,000 square miles. It contains
much valuable geological information; a study of the movement
of glaciers over the region; appendices on Chippewyan and
inland Eskimos vocabularies; and a third one on plants collected
ee J. W. Tyrrell. H. S. W.

. Livers of North America, a reading lesson for students of
ee aphy and geology ; by Israrn C. RUSSELL; pp. 1-327,
figures 1-28, plates i-xvii. New York, 1898 (G. P. Patnam’s
Sons). —Prof, Russell’s “Rivers of North America” is a fit com-
panion for “Lakes of North America,” “Glaciers of North
America” and ‘‘ Volcanoes of North America” by the same
author. It is a delightful “ reading lesson” for any one interested
in. geography or geology and will make brooks and rivers more
attractive to the lover of nature. The characteristics of stream
action are presented clearly in plain language without detailed
discussion of laws or theories, and each point is illustrated from
American rivers. The chapters on “Stream Development ” and
“The Life History of a River” trace the stream’s history from
youth to old age, tell of its birth, its efforts to adjust itself to its
environment and the accidents which it may suffer. The river is
made to appear as a thing of life. Students who have searched
through manuals, geological reports and scattered essays for facts
and laws of stream action will be thankful for the discussion of
these subjects here given. H. E. G.

9. Harth Sculpture or the origin of land forms ; by JAMES
GEIKIE; pp. 1-397, pl. i-ii, figs. 1-89. New York, 1898 (Giak:
Putnam’s Sons).—This volume is the latest number of The Science
Series published by the Putnams. The book treats first of the
agents of denudation and the resulting land forms in regions of
horizontal], inclined, folded and displaced strata. After this the
modifying effects of igneous, glacial and eolian action on surface
features are discussed. ‘‘ Coast Lines” and “ Classifications of
Land Forms” are very useful chapters. The book is addressed
to the average student who may desire some general knowledge
of the development of land forms, and we know of no other Eng-
lish work which gives a general account of the whole subject.
To American readers some of the terms and descriptions of
foreign localities will seem unfamiliar and the glossary in the
Appendix will be needed. This is particularly true of the divisions
of the Geological Time Scale.

As an introductory treatise Prof. Geikie’s book is very welcome
and students of nature who have mastered this will read Sir A.
Geikie’s ‘‘Scenery and Geology of Scotland,” Lubbock’s “ Scenery
of Switzerland,” Powell’s “ Canyons of the Colorado” and other
physiographie classics with added interest. H. E. G.

Geology and Mineralogy. 73

. Elemente der Gesteinslehve, von H. RosEnsuscn. 8vo, pp.
Ele Stuttgart, 1898.—This book is based, as the preface states,
on the lectures of the author at Heidelberg. While it claims to
be only an elementary treatise, it is much more than that, and will
be read with interest by every petrologist as giving the latest
views of its author and an excellent résumé of the present state of
the science from a certain, rather subjective, standpoint.

After an introductory chapter on the general principles of
petrography and the composition, structure and classification of
rocks, the author describes not only the igneous rocks, to which
his previous volumes have been devoted, but also the sedimentary
rocks and crystalline schists.

In the portion devoted to the igneous rocks, the essexites,
shonkinites and missourites have been raised to the rank of groups
of the same order as the granites. The monzonites, for which
Brogger has proposed a similar position, are left as a subdivision
of the syenites, while among the effusive rocks a group of trachy-
‘dolerites has been formed. These, according to Rosenbuscb, are
homologous with the essexites, but, though they resemble these
in being intermediate, the examples and analyses which are given
resemble more closely Brégger’s monzonites. The necessity for
the recognition of such groups of intermediate rocks is urgent,
but this group, as here given, is the least satisfactory of all,
embracing as it does many types which differ much among them-
selves, and it seems probable that it will undergo many changes
and subdivisions in the future. It may be suggested that the
name Jatite previously proposed by Ransome seems preferable to
the coextensive term of trachydolerite used by Rosenbusch.

The sedimentaries are very fully and satisfactorily treated and
the same is true of the metamorphic schists. For schists derived
from igneous rocks the author proposes the use of the prefix ortho,
and for those derived from sedimentaries the prefix para, so that
we would have orthogneiss, paragneiss, etc.

In general the views held by the author are the same as those
expressed in his previous works. The age distinction is practi-
cally banished, but traces of it still survive among the effusivcs.

The chemical characters of the various rocks are discussed in
considerable detail, and a valuable feature is the insertion of
tables of typical analyses, not only of the rocks themselves but of
their component minerals, many of which are published here for
the first time. They are in genera] well selected, but the almost
total lack of references throughout the book, especially for the
analyses, is to be deplored, even though such omissions are due
to the elementary character of the work.

But such criticisms are chiefly matters of personal opinion, and
the name of the author is sufficient guarantee for the high char-
acter of the volume. It should prove a valuable text-book for
the German-reading student, and will be a much used addition to
the library of every petrologist. H. 8. W.

74 Scientific Intelligence.

11. The Educational Series of tock Specimens collected and
distributed by the U. S. Geological Survey ; by J. S. Dituzr.
Bulletin No. 150, U. 8. Geol. Surv., Washington, 1898, pp. 400,
47 pl.; price 25 cts—A number of years ago the United States
Geological Survey undertook the praiseworthy task of collecting
and distributing to our leading educational institutions, as an aid
in geological teaching, a series of rock collections illustrating the
petrology of the United States. Under the able management of
Mr. J. S. Diller, aided by many geologists both of the Survey
staff and from other portions of the country, this task was
sucessfully completed, and about 250 sets, comprising over 150
specimens in each, were distributed. In addition, to supplement
the value of these collections, the Survey has now issued the
volume above mentioned, in which the rocks are described from
the standpoint of modern petrography. Under the editorship of
Mr. Diller, who also contributes largely to the descriptive
matter, the rocks have been described by a number of specialists,
in great part those by whom they were collected, and to whom
is due our knowledge of the geology of the regions in which they
occur.

The method of classification is simple and rational, and the
volume constitutes in fact an excellent practical text-book of
petrology for Americans, and as such will be found of very great
service to teachers of geology, as well as to those engaged in the
more specialized petrographical branches. Its appearance in con-
nection with the use of the collections cannot fail to give an
impulse to the study of petrography, and the Survey staff in
general, and Mr. Diller in particular, deserve great credit for
the completion of this useful and public-spirited work. L. v. P.

12. The Mechanical Composition of Wind Deposits ; by JOHAN
Auveust UppEN, pp. 1-69. Augustana Library Publications, No.
1. Rock Island, II]., 1898.—The author has carried on a long
series of observations in order to show the part played by the air
in motion in transporting sand deposits to different distances as
determined by their size. Four classes are especially recognized :
(1) lag gravels or coarse residue deposits in the rear of sand
dunes; (2) drifting sand, constituting the sand dunes of dry and
sandy regions ; (3): fine sand, soon dropped by the wind in the lee
of drifting dunes: (4) atmospheric dust, which only slowly settles
out of the air far away from the place where it was raised.

Many observations in regard to each of these have been made
at different points, chiefly in Illinois, Indiana, Nebraska, Kansas
and the Dakotas. The results, given in the form of numerous
tables, show the percentage of particles of characteristic size
present under the given conditions. For example, lag gravels,
though varying widely in different localities, include chiefly par-
ticles from 4™™ to 4™™ in diameter; drift sands those from 1™™ to
4™™; dune sand is more uniform, from 60 to 70 or 80 per cent of
the ‘particles ranging between 4 + and 4™™; lee sand runs down to
qis™™ or smaller, and atmospheric dust ' varies from $ to zi_™™ and

Geology and Mineralogy. 75

smaller. The approximate maximum distances over which quartz
fragments may be lifted by moderately strong winds in single leaps
are estimated as follows:

Gravel (8-1™™ diam.) A few feet.
Coarse and medium sand (1-4™") Several rods.
Fine sand (4-4™™) Less than a mile.
Very fine sand ({-;™”) A few miles.
Coarse dust Cae. 200 miles.
Medium dust (-~¢7") 1000 miles.

Fine dust (he ad less) Around the globe.

A striking result brought out is the comparative definiteness
with which the different grades are separated from each other by
the sorting action of the wind. In closing, the author discusses
briefly the relation of the facts brought out by his investigations
to the origin of the loess deposits, which in their mechanical com-
position certainly resemble atmospheric sediments. He does not
regard it possible, however, to reach a final conclusion from the
data at hand.

13. Brief notices of some recently described Minerals.—SENAITE
is a new mineral, related to ilmenite, described by Hussak and
Prior. It is found in rounded fragments and rough crystals in
the diamond-bearing sands of Diamantina, Minas Geraes, Brazil.
Crystallization like ilmenite, rhombohedral-tetartohedral | (tri-
rhombohedral). Hardness about 6, specific gravity of crystals,
5°30, massive 4°78 to 4:22; luster sub-metallic and color black, in
thin splinters, greenish. Analysis gave:

TiO. PbO FeO Fe.0; MnO MgO Sn02
G_.-4°78 57°21 10°51 4°14 20°22 7°00 0°49 0°11==99°68

The formula (Fe, Pb) O.2(Ti Mn) O, is suggested, but doubtful.—
Min. Mag., xii, 30, 1898.

MossiTE is a name given by Brogger to a nisbo-tantalate of
iron occurring with yttro tantalite at Moss, Norway. It is found
in black tetragonal crystals with c= 0-644. The crystals are
mostly twins with (101) as twinning plane; they are often dis-
torted by elongation parallel to (111) as is sometimes observed
with rutile. An analysis gave (N6,Ta),O, 82°92, FeO 16°62,
SnO, 0°18 = 99°72, which corresponds to the formula Fe (N6,
Ta),O,; further it was found that N6é: Ta=1:1. The author
shows that mossite is near tapiolite, which, however, is richer in
Ta,O, and has c= 0°652. He has also made the interesting
observation that the crystals of “tantalite” which have long
been figured in the text-books, are in fact identical with tapiolite,
being twin crystals similar to those mentioned above.— Vid. Skrift.
Math.-nat. Klasse, 1887.

VALLEITE is a mineral closely related to anthophyllite occur-
ring with the violet tremolite of Edwards, N. Y.; it is described
by G. Cesaro. Crystallization orthorhombic; occurs in colorless
prismatic crystals with the usual amphibole angles of 125° 30’
and 54° 30’. The cleavage is prismatic and pinacoidal. The
hardness is 4°5, specific gravity 2°88. Analysis gave :

76 Scientific Intelligence.

SiO. MgO CaO Fe,0; MnO K,O H.O

58°02 “27°99 5:04 1°28 2°88 0:89 3°13 == 99023
The formula is RSiO,, or that of anthophyllite, from which it is
stated to differ somewhat in optical characters (y — a= 0°0036).—
Leitschr. Kryst., Xxx, 84.

CxEDARITE is a fossil resin, resembling amber, from the alluvium
of the Saskatchewan river, Canada. Analysis gave: C 78°15,
H 9°89, S 0°31, O 11°20, ash 0°45 = 100. Described by R. Klebs,
Jahrb. ’ Min., li, 212 ref, 1898.

BaTAavitE, described by Weinschenk, is a decomposition pro-
duct occurring with the graphite deposits near Passau, Bavaria.
It occurs in white pearly scales with a specific gravity of 2°183.
Analysis gave: SiO, 42°33, Al,O, 16°35, MgO 28:17, H,O 13°19 =
100:04. Jfor this the composition suggested is H, Mg, Al, Si, O,,.—
Leitschr. Kryst., xXviil, 160.

GRUNLINGITE, described by Muthmann aad Schroder, is a
sulpho-telluride of bismuth, related to tetradymite, from Cumber-
land, England. It was earlier investigated by Rammelsberg,
but the authors have obtained on analysis a composition agreeing
with the formula Bi,S,Te or Bi(S, Te). It appears in cleavable
masses with metallic luster and gray color; the specific gravity
is 7°321. Named after Dr. F. Griinling of Munich.—Jdid., xxix,
144,

PLANOFERRITE, described by L. Darapsky, is a hydrated ferric
sulphate from the Lautaro copper mine in Atacama. It occurs
in tabular crystals which are colorless or lemon-yellow. An
analysis gave: SO, 15°57, Fe,O, 31:20, H,O 51°82, insol. 1°41 =
100. This corresponds to Fe,O,.SO,.15 H,O.—TZbid., xxix, 213.

IJ. Botany anp Zoouoqy.

1. The Potsonous effect exerted on living plants by Phenols.
(Botan. Centralblatt, Nov. 15, 1898.)—Professor R. H. Truz
has extended his fruitful investigations respecting the action of
toxic substances far beyond the boundaries which he originally
marked out. With efficient collaborators, he studied two years
ago the effects of salts, acids, and bases, on one of the Legumi-
nose, and had the satisfaction of having the results substantially
confirmed by other experimenters, and the suggestions carried
into a wider field. With Mr. Hunkel, he returns to the task, and
obtains results which appear to confirm the belief that the whole
subject of Toxicology is to receive abundant light from the study
of plants. It is well known that the meagre information regard-
ing the action of active remedial agents on plants, embodied in
the Dispensatories, is unsatisfactory, but it certainly indicates a
desire on the part of the commentators to get what little hight
glimmers through the obscure experimentation. Now, of late
years, the whole subject has been placed on a different basis, and
the work done upon this basis is creditable alike in its suggestion
and its execution. It is not yet time to summarize the details or

Botany and Zoology. 6

attempt their complete co-ordination, but it is not too soon to
predict that materials will before long be sufficient for a rational
discussion of many obscure points in Toxicology and, we dare to
hope, in Therapeutics. Tothe energy and carefulness with which
Dr. True has prosecuted his work thus far, we owe most import-
ant contributions, and we trust he will be encouraged to carry
his researches still further afield. The new paths which he has
been instrumental in opening up are likely to attract many
explorers. It is to be sincerely wished that all who enter the
_ new field will bring to the work the same acuteness, patience, and
thoroughness which seem to characterize the work thus far. To
those of our readers who may like to know the thought which
inspires the work, we may use words of Dr. True, which are
abundantly justified. ‘The application of the theory of dissocia-
tion of electrolytes.to explain the toxic action of acids, bases, and
salts, on living organisms has yielded results of greatest import to
both chemistry and biology.” Now. beyond these substances
others are to be investigated by the same methods: let us hope
with equally satisfactory results. G. L. G.

2. On the peculiar mode of formation of pollen-grains in
Magnolias.—In the number of Comptes Rendus for Oct. 24, 1898,
GUIGNARD gives an extremely interesting account of one of the
puzzling exceptions to general rules, which are full of suggestions
as to classification of organisms.

The author calls attention to the general characters of pollen
formation which serve to separate more or less completely mono-
cotyledons from dicotyledons. In the former, the first division of
the nucleus of the mother-cell is followed by partition of the lat-
ter, and then the two daughter-cells divide in their turn, in the
same way. In dicotyledons, on the other hand, the first nuclear
division is not followed by partition of the mother-cell; that does
not take place until after the second nuclear division, between the
four nuclei which it has furnished.

Now to this general rule, Guignard had already pointed
out a remarkable exception, namely, that among the monocoty-
ledons are to be found plants which behave in this respect
just as do the dicotyledons. These exceptions are the Orchid-
acew. They nearly all undergo simultaneous quadripartition of
the pollinic mother cell. The author now announces that in the
Magnolias is to be seen a type of division unlike either the
first or the second. The type is almost intermediate between
the two hitherto recognized, but by the formation of an incom-
plete partition (possibly sometimes complete) immediately after
the first nuclear bipartition, is more like that of the monocotyle-
dons than of the dicotyledons. G. L. G.

3. Hlements de Botanique. P. Van Tizaunem. 2 volumes,
16mo. Paris, 1898.—In this revised edition of a helpful work,
Professor Van Tieghem gives his own views regarding certain
disputed points relative to structure, and states his reasons for
suggesting a new classification of the higher plants. The princi-

78 Scientific Intellagence.

pal features of this new system have already been presented in
this Journal, and therefore need not be again spoken of in detail.
But it seems proper to mention the fact that Professor Van
Tieghem’s system, although distinctly revolutionary, appears to
offer advantages in practical ways, especially in bringing out
clearly the disguised resemblances. ‘The author has done good
service in emphasizing a good many things which have been more
or less overlooked in previous systems. It is not improbable that
some of these things have been insisted on rather too strongly, and
in this way, may throw the system somewhat out of proportion,
but thus far, this does not seem to be the case. At any rate, the
sense of general proportion is good, and, as a natural result, the
perspective is not misleading. ‘Teachers will find a good many
fruitful suggestions in every part of the two volumes, but most of
al] in the portion devoted to classification. G. L. G.
4. Sketch of the Evolution of our Native Fruits; by L. H.
BalLEY; pp. xiii+472, 125 figures in text. New York, 1898 (The
Macmillan Co.).—During the past ten years Professor Bailey has
made a study of our indigenous fruits and of the cultivated vari-
eties to which they have given rise. Some of the results of his
investigations have already appeared in periodicals, and the
present volume brings these scattered observations together and
presents the whole subject in a more complete and connected
form. The greater part of the book deals with the more important
of our fruits, those which are most widely cultivated and which
are consequently represented by the greatest number of cultivated
varieties. These include the grape, the mulberry, our native
plums, cherries and apples, the raspberry, the blackberry and the
dewberry. The chapter on the grape, though particularly
exhaustive, may indicate the scope of the work: in addition to a
history and description of our various native grapes and of their
cultivated varieties, it gives an account of the early attempts and
failures to introduce the European grape into eastern America
and concludes with a full bibliography of American grape litera-
ture. At the close of the book some attention is paid to our less
important fruits, the gooseberries, currants, etc., which, although
less widely cultivated, offer considerable promise to the cultivator.
A. W. EB,
5. Bush-Fruits: a Horticultural Monograph of Raspberries,
Blackberries, Dewberries, Currants, Gooseberries, and other Shrub-
like Fruits ; by Frep. W. Carp; pp. xii+337, 113 figures. New
York, 1898 (The Macmillan Co.).—The present book is the first
of a proposed series of monographs dealing with the various types
of American fruits. As its title indicates, it is distinctly horti-
cultural and devotes its chief attention to a description of the
varieties of the “ bush-fruits” enumerated and of the methods
employed in their cultivation. A. W. E.
6. Lhe Metamorphosis of Asterias pallida [A. vulgaris| with
special reference to the fate of the Body Cavities ; by 8. Goro.
Jour. College Science Imp. Univ. Tokyo, Japan, vol. x, pt. il,

Botany-and Loology. 79

pp. 239-278, pl. xix—xxiv, 1898.—This important paper relates to
the common northern starfish of New England. It is the result
of work done at the Newport laboratory of Mr. A. Agassiz and
at the Museum of Comp. Zoology of Harvard University. _ v.

7. Zoological Results based on material from New Britain,
New Guinea, Loyalty Islands, and elsewhere, collected during
1895-1897; by ArTHuR Wittrey. Part I. Cambridge, Eng.,
Univ. Press.—The first article consists of a very detailed account
of the development and anatomy of Peripatus Nove-Britannie,
sp. nov., illustrated by four plates. The second is devoted toa
new species of Caprellidz (DMetaprotella Sandalensis). The third
is on a sea-spake from the South Pacific, Atpysurus annulatus
(Krafft), with figure. The fourth is a Report on the Centipedes
and Millipedes, pl. vi. The fifth is on the Phasmide with notes
on the eggs, pl. vii-ix. The sixth paper relates to the aoe
and Scorpions, pl. x, xi.

8. An Account of the Crustacea of Norway, with figures of all
the species ; by G. O. Sars’ Vol. II, parts xi, xii. Oniscide,
Bopyride, Dajide; 4to. Bergen. Published by the Bergen
Museum, 1898. This is a continuation of the admirable work of
Professor Sars on the N orwegian Crustacea. This part contains
eight autographic plates drawn by Dr. Sars himself, with his
customary care and accuracy. This work is of special interest to
American naturalists, because a considerable percentage of bee
species will be found ‘also on our coast.

9. The Mollusca of the Chicago Area: The Pelecypoda ; ee
Frank Cotuins BAKER, pp. 1-130, with 27 plates. The Chicago
Academy of Sciences, Bulletin No. ill, Part i of the Natural
History Survey, Sept. 1, 1898.—This exhaustive and liberally
illustrated memoir by Mr. Baker has recently appeared. A
second part, devoted to the Gastropoda, is promised.

10. Catalogus Mammalium tam viventium quam fossilium a
Doctore E.-L. Troverssart, Parisiis. Nova Editio (Prima com-
pleta). Fasciculus IV, pp. 665-998; V, pp. 999-1264. Berlin,
1898 (R. Friedlander & Son).— Parts 1-11 of the exhaustive Bib-
liography of papers relating to living and fossil mammals, by Dr.
Trouessart, have already been announced in our pages. Parts iv
and vy have now appeared, embracing some six hundred pages
and bringing this important work toa close. Part iv includes the
Tillodontia and Ungulata: Part v the Sirenia, Cetacea, Edentata,
Marsupialia, Allotheria, and Monotremata. The genera included
in these two parts extend from Nos. 761 to 1588, and the species
from Nos. 4086 to 7224; and besides there are many varieties not
separately numbered. The whole bibliography is printed with
admirable clearness.

1l. The Fishes of North and Middle America; by Davin
STARR JORDAN and Barron WARREN EVERMANN; pp. 1-xxx,
1241-2183. Washington, 1898 (Bulletin 47 of the U.S. National
Museum, Smithsonian Institution).—Part II of the complete
descriptive catalogue of the fishes found in the waters of North

80 Scientific Intelligence.

America north of the Isthmus of Panama, by Jordan and Ever-
mann, has recently been issued. Part I was published in October,
1896. The third and final part with index, glossary, etc., is prom-
ised soon; it will be followed by a fourth volume which will be an.
atlas of plates. : .

TV. MIscELLANEOUS SCIENTIFIC INTELLIGENCE.

1. Institute of France.—At the session of the Académie des
Sciences held at Paris, December 5, 1898, Professor O. C. Marsh
of Yale University, to whom the Cuvier prize was awarded a
year since, was elected Correspondent of the Academy.

2. Annual Report of the Board of Regents of the Smithsonian.
Institution showing the operations, expenditures and condition of
the Institution to July, 1896; pp. i-li, 1-727; plates i to lxi.—
This valuable report submitted to Congress by the Secretary,
S. P. Langley, has recently appeared. The opening portion (pp.
1-77) are given to matters of administrative detail in regard to-

the work of the Institution in its different departments. The
General Appendix (pp. 89-727) contains as usual a series of sci-.
entific papers on a wide range of topics, the republication of
which in this place is a boon to many interested in science.

38. Der Ursprung der Afrikanischen Kultur von L. FROBENTUS,.
pp. 1-xxxi, 1-368, with 5 folded charts, 9 plates and 225 illustra-
tions in the text. Berlin, 1898 (Gebriider Borntraeger).—This is-
the opening volume of what promises to be an extensive and
important work, devoted to the origin of culture in general. The
subject of the volume is the origin of culture in Africa, and this
is treated with much fullness and abundance of illustration. Some
of the topics discussed in detail are the weapons, shields, knives,
etc.; also the various kinds of musical instruments, the different.
forms of huts and household utensils. The latter portion of the
work deals with the physiological side of the subject in its varied
aspects.

4. Organic Evolution Considered; by ALFRED FATRHURST,
pp. 1-386. Christian Pablishing Co., St. Louis, 1897.—In regard
to his work the author states: “I look upon the theory of evo-
lution as being of no importance except as it involves the well-
being of man. . . My object in what I have written is to promote
the belief in theism and in the existence of a spiritual nature in
man which theism alone can explain.”

Catalogues recently issued.

Catalogue of Books and Papers on Chemistry and Physics, No.iv. Dulau and Co.,
37 Soho Square, London, W., 1898.

Dr. F. Krantz, Rheinisches Mineralien-Contor in Bonn.—Katalog Nr. la
(Siebente Auflage), Mineralien and Mineralpraparate, Mineralogische Apparate
und Utensilien, pp. i-xii, 1-113, 1898.

Catalogue and Price List of Minerals for Scientific and Educational Purposes.
Loose crystals a specialty. Roy Hopping, 5 and 7 Dey Street, New York City.
1899.

_ OnJanuary 3d we will commence moving to our new
store, 812 and 814 Greenwich Street. Our 12th Street
store will be our address until Tuesday, January 17th,
when we will open in our new location. Our new
quarters will give us more than double the space we.

_ now have, far better light, and all the conveniences re

. of a modern building. We believe the location will a
be found even more convenient than where we are Beer

to the majority of our customers. Fuller particulars
will be given in a Winter Bulletin which will be
mailed to customers early in January.

NEW MINERALS IN OUR NEW STORE.

We have been holding back all new arrivals for several weeks past in order
ab have a large amount of brand new stock to offer at our

OPENING SALE, TUESDAY, JANUARY 17.

i= grichAlcite from Montana. A splendid large lot of specimens superior
to any ever before found, 2d5c. to $3.90.
_Cerussite from Missouri. The best crystallized specimens ever discovered
in the U.S. Nineteen boxes. Large and small. Many associated with Hae
bright green Pyromorphite. The entire output of the little cave secured Roe s
by us and will be placed on sale January 17th. 10ce. to $5.00. ee
_ Sphalerite from Missouri. From a new locality we have secured a small
lot of the most lustrous, beautiful and perfectly crystallized Missouri Spha- |
_ lerites we have ever seen, 25c. to $2.50. Pas
_ Missouri Marcasite. Several hundred sharper, brighter, better Marca- eee
sites than ever before. 10c. to 50c. a
_. Other new Missouri Minerals. Very attractive Galenas from a new .
locality ; 10c. to $1.00. A few extra choice Caleites, highly modified, clear,
beautiful golden color.
_ Wavellites from Arkansas. A wonderfully fine, large eee
bly superior to any we have ever had before. 10c. to : 32. 00. Also a large :
lot of choice Variscite from Arkansas. 10c. to $1. 50. Rab -
_ Crystallized Cinnabar. Over 100 choice specimens. 25c. to $5.00. ; eae
Yellow Wulfenite. A new find of groups of extra large crystals. 25c. ees,
to $6.00. aie
Colorado Thomsonites. A very large lot of magnificent specimens, 10c. Vinee
‘to $3.50. Colorado Analcites, choice groups, 10¢, to $1.50. Bese
Unaltered Crocidolite. A new lot just received. Large, firm specimens Gh em
of long fibre, beautiful chatoyance and the peculiar blue color. 10c. to 50c. kaa
4S apanese Stibnites. A little lot of excellent, small terminated crystals. rae
— d0c. to $2.50. 17
_ Japanese Twin Quartz. The largest and best ever imported. Only a ‘VANE
- few of them. $75.00, $50.00, 330.00, $12.50 down to $5.00. Also a few i
small twins remaining from our old stock at 35¢. to $1.00. Dye
_ MANY OTHER CHOICE SPECIMENS will be placed on sale at our _ Gee
pe opening in our new store. mee

GEO. L. ENGLISH & CO., Mineralogists,

q is 64 East 12th Street (until Jan. 16), 812 and 814 Greenwich Street
8 (on and after Jan. 17), New York.

CONTENTS.

Page
Art. L_—Thermodynamic Relations of Hydrated Glass; by a

OSMBARUS 2005 ee ot Re ee ete AA ee
IJ.—Platinum and Iridium in Meteoric Tron; Py Jee |

SPAVISON 0 oe ee a ee
IIT:—Studies in the Cyperacee; by T. Hoim____._.-_.-=2 B/D:
IV.—Regnault’s Calorie and our Knowledge of the Specific

Volumes of Steam; by G. P. StaRK WEATHER -_-___-_- 13
V.—Estimation of Borie Acid; by F. A. Goocuw and L. C. ©

JONES 2222 ek 2a LL aL ee 34
VI.—Descriptions of imperfectly known and new Actiniauns

with critical notes on other species, IL; by A. E. Vur-

ST ene oks ee ty LO SRA ee 41
VII.—Mineralogical Notes; by W. F. Hittepranp~-_----- 51
VIII.—W hat is the Loess? by F. W. SarpEson-_--_----- 58
IX.—Absorption of Gases in a High Vacuum; by C. C, ©

HUTCHINS ©}. oon ccs oR ees eee

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Extraction of Nickel by the Mond Process, RoBERTs-
AUSTEN: Atherion, Crookes, 64.—Preparation of Graphitie Acid, STAUDEN-
MATER: Compounds of Lithium and Calcium with Ammonium, MoIssan, 65,— —
Spectra of Iodine, KoNEN: Manual of Chemical Analysis, Qualitative and
Quantitative, G. 8. NewrH: Stratified Brush Discharges in Atmospheric Air,
TorPLer, 67.—Spectrum of Lightning: Dispersion in the Hlectriecal Spectrum,
F. Marx: Traité Elémentaire de Méchanique Chimique fondée sur la Thermo-
dynamique, P. DuneM, 68.—Prismatic and Diffraction Spectra; Memoirs of
Joseph von + raunhofer, 69.

Geology and Mineralogy—Maryland Geological Survey, 69.—Lower Cretaceous
Grypheeas of the Texas Region, R. T. Ht and T. W. VauGHan: Bibliographie
Index of North American Carboniferous Invertebrates, S. WELLER, 70.—Con-
tributions to the Tertiary Fauna of Fiorida, W. H. DAtL: Contributions to
Canadian Paleontology, J. F. WuHITEAVES: Geological Survey of Canada: Re-
port on the Doobaunt, Kazan and Ferguson Rivers and the Northwest Coast of
Hudson Bay, J. B. TyRRELL, '71.—Rivers of North America, a reading lesson
for students of geography and geology, I. ©. Russe~u: Earth Seulpture or the
origin of land forms, J. GEIKIE, 72.—Elemente der Gesteinslehre, H. ROSEN-
BUSCH, 73.—Educational Series of Rock Specimens collected and distributed by
the U.S. Geological Survey, J. S Dimer: Mechanical Composition of Wind
Deposits, J. A. UDDEN, 74.—Brief notices of some recently described Minerals, -
75.

Botany and Zoology—Poisonous effect exerted on living plants by Phenol R. H.
TruE.—Peculiar mode of formation of pollen-grains in Magnolias, GUIGNARD:
Elements de -Botanique, P. Van TrzGHEM, 77.—Sketch of the Evolution of our
Native Fruits, L. H. Baitey: Bush-Fruits, F. W. Carp: Metamorphosis of |
Asterias pallida with special reference to the fate of the Body Cavities, 8. Goro,
78.—Zoological Results based on material from New Britain, New Guinea,
Loyalty Islands, and elsewhere, collected during 1895-97, A. WiLLEY: Account:
of the Crustacea of Norway, with figures of all the species, G. O. Sars: Mol-
lusca of the Chicago Area: The Pelecypoda,-F. C, Baker: Catalocus Mamma-
lium tam viventium quam fossilium, E.-L. TRouESsART: Fishes of North and
Middle America, D. S. JornDAN and B. W. EVEeRMANN, 79.

Miscellaneous Scientific Intelligence—Institute of France: Annual Report of the

Board of Regents of the Smithsonian Institution: Ursprung der Afrikanischen
Kultur, L. FROBENIUS: Organic Evolution Considered, FAirHuRsT, 80.

AMERICAN
JOURNAL OF SCIENCE.

EDWARD 8. DANA.

EDITOR:

ASSOCIATE EDITORS

Prorussors GEO. L. GOODALE, JOHN TROWBRIDGE, »
; H. P. BOWDITCH ano W. G. FARLOW, or Camprincs,

Prorsssons Gy C. MARSH, A. E. VERRILL anv H. S.
WILLIAMS, or New Haven,

- GEORGE F. BARKER, or Putapetpura,
_ Prorrssor H. A. ROWLAND, or Barrimore,
- Mr. J. S. DILLER, or Wasurneron.

FOURTH SERIES.
VOL. VII-[WHOLE NUMBER, CLVII.]

No. 38.—FEBRUARY, 1899.

WITH PLATE I.

NEW HAVEN, CONNECTICUT.
1899.

PRINTERS, 125 TEMPLE STREET.

TUTTLE, MOREHOUSE & TAYLOR,

blished monthly. Six dollars per year (postage prepaid). $6.40 to
n subscribers of countries in the Postal Union. Remittances should
de tence by money orders, registered letters, or bank checks.

Le oP ie Pe a

JUST OUWm
February,.1899. |

MAILED FREE ON APPLICATION.
ILLUSTRATED SUPPLEMENT i

TO THE ‘‘ COMPLETE MINERAL CATALOGUE” AND ~

ANNOUNCEMENT OF NEW ARRIVALS.

Sixteen pages. . Two handsome photo-engravings of the beautiful stellate and
‘‘spear-head ”’ twin Cerussites, and Anglesite coating same. Brief descriptions —
are given of important accessions to stock. Among them are the wonderful
erystallizations purchased by our Australian collector during his five weeks’ trip
to Broken Hill, which is becoming known as one of the greatest of mineral locali- -
ties. Gem-like Raspites and Stolzites, Twin Cerussites, brilliant honey-colored
Anglesites, Strontianocalcites, Iodyrites, Pyromorphites, ‘etic. Gorgeous Tasma-
nian Crocoites, Franklin minerals (three new species), Endlichites - = other
great finds) REVISED POUND LIST.

THE OLD TRAUTWINE COLLECTION is undergoing rapid disinte-
gration before the stream of orders directed against it. Collectors were not slow
to accept the invitation extended a month ago. Their desiderata lists are being
cut down in a most satisfying way, through the purchase of Trautwine specimens.
A full list of the more important is given in the Supplement. If you want long
forgotten and obscure rarities as well as choice examples of commoner species
found in the early part of the century, write at once. They will be gone in a
few weeks.

THE: “NEW ‘SPECIES? tia

one of the important features of our ‘‘ Complete Mineral Catalogue,” is brought
up to date in this supplement. Thus you secure free, a careful review, compiled ©
at considerable expense, of mineralogical discoveries between September, *97, and
February, 99. It gives name, reference to original publication, form and com-
position of all new minerals. (The ‘‘Complete Mineral Catalogue,” 186 pp., 40
illustrations, with synopsis of Dana’s classification, name, composition and form
of every known species; supplement to September, ’97, alphabetical index, metal-
lic classification and other useful tables with Price Lists. Paper bound, 25c.;
cloth, 50c. A work of reference invaluable to mineralogists.) February Supple-
ment is free whether catalogue is ordered or not.

ALL PRICE LISTS FREE.

Dr. A. EK. FOOTE,

WARREN M. FOOTE, Manager.

Mineral Specimens for Scientific and Educational Purposes.

1317 Arch Street, Philadelphia, Pa., U. 8. A.

Established 1876.

THE

AMERICAN JOURNAL OF SCIENCE

[FOURTH SERIES. ]

————++e—______

7 oe X.—A Contribution to the Study of Contact Meta-

morphism ; by J. MoRGAN CLEMENTS.
_[Published by permission of the Director of the U.S. Geological Survey.]

THE various Huronian sediments which form a great por-
tion of the iron-bearing districts of the Upper Peninsula of
Michigan have in all of these districts been found to be pene-
trated by dikes of igneous rocks, which are predominately
basic incharacter. The influence which these dikes exert in
the formation of ore bodies where they cut the iron-bearing
formations has been carefully worked out and described by
Van Hise in numerous papers. The effects of contact meta-
morphism produced by the dikes upon the intruded sediments
are of much less economic importance, and, to a very consider-
able extent, have been overlooked. In this paper I shall
describe the products which have resulted from the intrusion
of basic dikes in the Mansfield slate formation, a Lower
Huronian iron-bearing formation of the Crystal Falls district
of the Upper Peninsula of Michigan.*

The Mansfield formation, in its typical development, occu-
pies a narrow valley, about three miles in length, through
which flows the Michigamme River. In this valley there is
situated the village of Mansfield and the mine of the same
name. The formation consists of graywackes, slates (these
predominate, hence the name), and phyllites, of demonstrably
sedimentary origin. Associated with these, either as beds or

* For details concerning the position which the Mansfield formation holds with

reference to the other Huronian rocks, the reader is referred to an article entitled

The Crystal Falls Iron-bearing District of Michigan, soon to be published in the
19th Ann. Rept. U. S. Geol. Survey, and in a Monograph of the Survey.

Am. Jour. So1.—Fourts Series, Vou. VII, No. 38.—Fepruary, 1899.
6

82. SJ. M. Clements—Study of Contact Metamorphism.

as lenses agreeing in strike with the sedimentaries, we find
slaty siderite, ferruginous chert and iron ore. The origin of
the siderite and chert cannot be stated with certainty. The
ore was derived from them by processes of metasomatism and
concentration. All of these rocks, wherever exposed, are
found to strike north and south, with at times slight variations
from this direction, probably due to minor crumpling of the
formation. The dip is high to the west at about 80°.

Overlying the Mansfield slates is the voleanie Hemlock
formation.

Immediately east of the slates, that is, stratigraphically
under them, there occur continuous masses of coarse dolerite
(diabase), which cut off the slates in their strike, both to the
north and to the south. No well-characterized dikes of dolerite
cutting the Mansfield formation have thus far been found,
either by surface studies or by underground exploration. In no
instance, moreover, has adirect contact between the dolerites and
the slate formation been observed, owing to the contacts being
covered by disintegration products, although in many instances
these rocks are separated from each other by spaces only a
few feet in width. That the relations between the two, how-
ever, are those of igneous intrusion, is shown by the fact that
the dolerites cut across the strike of the slates. Furthermore,
if we make a traverse from the dolerites westward into the
slates we notice the following. . On the westward flank of the
dolerite ridge we find associated with the dolerites masses of
hard, peculiar, hornstone-like rocks, which have a well-banded
character and in places are of large size and very numerous.
Leaving this zone we reach the valley proper, in which lie the
normal sedimentaries of the Mansfield formation. It thus
appears that we have an intermediate zone between the doler-
ites and the slates. The banded rocks forming, with the asso-
ciated dolerites, this intermediate zone, are believed to be
sedimentary rocks which have been metamorphosed by the
dolerites.

The exact original characters of these sedimentary rocks
cannot be stated with positiveness. They are believed to have
been slates having essentially the same general physical charac-
ters and mineral and chemical composition as the slates which
now lie next to the contact zone and extend away from it for
several hundred feet across the strike, without showing any
noticeable changes in character. Although original chemiéal
variations exist in these slates, as is shown by the banding, and
did exist in the slates from which the metamorphosed products
to be described were derived, as is shown by the banding in.
them,—nevertheless, such initial variations in these rocks is of a
quantitative rather than qualitative character, and it is not

J. M. Clements—Study of Contact Metamorphism. 88

believed that they can vitiate the conclusion concerning the
effect of the metamorphic action upon the slates.

In the further discussion, therefore, it is assumed that the note-
worthy differences between the rocks to be described are due to
metamorphic action, and are not due to the original chemical
differences. This assumption seems to be warranted by the fact
that the changes are in perfect accord with those described
from similar dolerite-slate contact zones in other areas.

The products of the contact metamorphism are those banded
and spotted rocks which have been called spilosites, desmosites,
and adinoles, and they will be the special subject of this article.

The dolerites (diabases) are coarse-grained, and do not exhibit
anything of especial interest, consequently no description will
be given of them. No evidence of any endomorphie action
whatever has been observed in them.

In a study of contact action it is, of course, of the greatest
importance to be able to determine accurately the order of suc-
cession from the unmetamorphosed to the most metamorphosed
forms of the rocks. In the present case it is impossible to com-
pare sections from, or analyses of, any given stratum of the rock at
successive stages of approach to the line of contact, for the reason
that, where the exposures were found, the dolerite had been in-
truded parallel to thestrike of theslates. Moreover, the exposures
are very poor, indeed. This order of succession has, however,
‘been made out so satisfactorily for other localities by Lossen
and others, and the characters of each rock in the succession
have been so well described, that I have no hesitation, after
microscopical study of thin sections of the specimens, in pre:
senting the series in the following order. Beginning with the
clay-slate, the least metamorphosed rock in the district, and
the ones farthest removed from the intrusives, we pass to the
phyllites, then to the spilosites and desmosites, and finally to
those which are known as adinoles, the latter being those which
occur next to the intrusive.

The Clay-slates.

These are dull and lusterless banded rocks, ranging in color
from black to olive-green and red. They are usually impreg-
nated with more or less iron pyrites in large macroscopical
crystals. By high power, one can very readily distinguish in
the slates round or oval areas of limpid quartz, surrounded by a
dark grayish mass, which consists of minute flakes of white
mica, crystals of rutile, and some of hematite, here and there a
long, transversely-fractured, greenish needle, ‘taken for actino-
lite, and, lastly, a dark grayish to black oranular ageregate,
which is the chief coloring matter of the slate, and which
makes up a very considerable part of the rock. This aggre-

84. JS. M. Clements—Study of Contact Metamorphism.

gate is presumed to be of clastic origin, and to be chiefly feld-
ae dust, darkened by carbonaceous and ferruginous specks.

areful search was made for flakes of chlorite and biotite, but
no traces of either mineral could be found. Likewise, feldspar
is absent in grains or crystals of a size which could be recog-
nized. The banding of the slates is occasioned by the varying
quantities of these minerals, although each band contains some
individuals of all the kinds enumerated. All the crystals lie
with their long axes parallel, causing the cleavage of the slate.

t is clear that the slate is already in what we may perhaps call
a semicrystalline state, the crystals of white mica, actinolite
and rutile evidently being new developments in the sediments.

Origin and composition of the clay-slates.—The origin of
the clay-slates of the Mansfield formation is probably to be
looked for in the disintegration and decay of the Archean
ue and the subsequent compression of the resulting clay ;
or between the Archean granites and the Mansfield slates no
other rock masses are known to have existed in this region
from which the clay could have been derived.

An examination of series of analyses of granites shows that
while the relative percentages of soda and potash vary con-
siderably, on the whole the potash is the higher of the two.

As a consequence of the easier solubility of the soda, this
relation between the alkalies, soda and potash, is maintained,
and is often very striking in clay-slates. An average of thirty-
one analyses of clay-slates taken from various sources shows
two and one-half times as much potash as soda. In the case of
the Mansfield slate, it will be seen that the difference is much
greater, there being present ten times as much potash as soda.

ANALYSIS OF MANSFIELD SLATE.
(By Mr. Geo. Steiger, U. S. G. 8S.)

STO) 2 ONS OL oe eee ee aes 60°28
TiO 2 TRE One eee Ons Lara 69
ANOS eG se aig Soe ae 22°61
Be,O) oo oo eke eee aie Gh oe
BeO 220 2) Re eS a ee "45
‘Min O iis Ge i ee ech ee trace
CaQie occ oe ee eee ae 13
BaQyt 5c eee ene eee "04
MBO. oi. aes ha eee ao 1°35
HG.) yw = oe See
ET 6 Rema LRP SE) uy BPE "b4
GO) at 100 = eee nee 60
TO, af 100 4. eee 3°62
ON oD Spo ae a a2 eet ae OE 03
Oe. seat al Seats et enn ne ae none
COLE el. OA eerie oe 97

Totalexn 222s oon

J. M. Clements—Study of Contact Metamorphism. 85

The relative proportion of the alkaline earths, lime and
magnesia, is also striking, the latter being present in the
greater quantity. As a general rule, in igneous rocks (and to
igneous rocks all clay-slates owe their ultimate origin) the
reverse condition exists, i. e., the magnesia is subordinate in
quantity to the lime, except in the ultra basic non-feldspathic
kinds. The carbon present in this slate is considered as offer-
ing trustworthy evidence of the presence of organic life at the
time of the deposit of the slates, though no more satisfactory
evidence of the existence of life thus early has ever been
found. It will be noticed finally that considerable water is
present, but in consideration of the character of the rock this
is to be expected, and if anything the value is low. These
clay-slates are the rocks which are nearest the original sedi-
ments.

Phylilites.

The phyllites have a silky luster and bluish black color.
They are composed essentially of white mica, which occurs in
large quantity surrounding grains of quartz and feldspar (?).
Rutile occurs in single crystals and in clumps of small crystals
scattered through the mica. Associated with it, there occur
here and there flakes of hematite, and small aggregates of
black, undeterminable specks. Apparently, no interstitial
material, such as occurs in the clay slates, is here present.

These rocks seem to differ from the clay slates only in that
they are more completely crystalline, the interstitial material of
the slates having disappeared. Owing to the apparently unim-
portant differences which exist between the clay-slates and the
phyllites, no analyses have been obtained of the latter.

Spilosites, Desmosites, and Adinoles.

With these rocks we begin the consideration of the true
contact products of the dolerite. These contact rocks possess
certain characters in common. They are dense, flinty, “ horn-
stone-like” rocks, which in some eases still show the fine
banding of the original slates. Others are very characteristic-
ally spotted. They have a splintery, and at times almost con-
choidal fracture, and vary in color on fresh fracture from light
to very dark gray and greenish. The weathered surface, in
almost all cases, is covered by a thin, white to light yellowish
erust. The mineralogical components are quartz, feldspar
(albite), biotite, chlorite, white mica, actinolite, rutile, epidote,

86 SJ. M. Clements—Study of Contact Metamorphism.

and iron oxide. Various combinations of these minerals
occur, and likewise the textures vary. As a result of these
variations there are produced the different kinds of contact
products known as spilosites, desmosites, and adinoles. The
characters of the minerals showing nothing unusual, I shall
briefly describe the textures and mineral combinations which
characterize these various rocks.

Spilosites—The ordinary spilosites are distinctly mottled in
the hand specimen, and show clearly to the naked eye in thin
section the oval spots which characterize them. These oval
areas are very commonly four millimeters long, and in rare
cases even longer. They are frequently connected, forming
chains. The spots are very appreciably darker than "the mass
in which they lie, and are composed of aggregates of chlorite,.
quartz, feldspar, and rutile, with a small amount of white mica.
The chlorite is the chief constituent of the spots and gives
them their dark color. The surrounding mass consists essen-
tially of white mica, quartz and feldspar, small epidote and
rutile crystals, flakes of hematite, and with a very slight amount
of chlorite. The different proportions of chlorite and musco-
vite cause the difference between the spots and the ground-
mass. In some of the spilosites we find a few flakes of biotite
and needles of actinolite. However, these are always very
subordinate in quantity to the chlorite.

Other spilosites have been noted in this area in which the
spots are white and lie in a fine-grained dark mass composing
the greater part of the slides. So far as I can learn, only one
similar instance of the occurrence of such a variety of spilosite
has been described. This is by Van Werveke, to whose de-
scription reference is made by Zirkel* and Rosenbusch.f

The white spots are composed essentially of feldspar, with
only a minor amount of chlorite and epidote. The feldspar
grains are much larger than those which take part in the con-
stitution of the mass surrounding the spots. This surrounding
mass is made up of quartz, feldspar, chlorite, epidote, some
sphene, with sheaves of actinolite scattered through it. In one
section flakes of biotite were observed mixed with the chlorite,
though in very subordinate quantity.

* Zirkel. Lehrbuch der Petrographie, 2d edition, vol. ii, 1894, p. 719.
+ Rosenbusch, Mikroskopische Physiographie, 3d edition, vol. ii, 1896, p. 1127.

J. M. Clements—Study of Contact Metamorphism. 87

ANALYSIS OF SPILOSITEs.

iv 2.
tetera ie) ove. Si Saeed 50°79
<5 RRS Sekt ry neg ee 1-70 "92
Al,O, Ye eens fae ee 8 19°00 19°35
SS ee eee none none
ge so, ee 1°29
TEL) AGS 2 cole ae ree 3°37
Je he eee trace trace
SE aes 1°55 1°71
0 ie be aes trace none
Nea eS oS. ee trace
it) eer 3:29 4°35
ee Se ‘70 22
Rien? 6°72 8:29
a eS trace none
H,O at » 2 5 | Se eae pg liane? os "34 "18
H,O amove EIQ" 2022272 3°26 2°34
ee 8 ee “15 “04
cae 0 Se a a ne none none
Rete OO. oo ee none
ones +S pee tae
ete eT ee oe none none
eee eeenee none

Maite 5S 99°72 99°76

No. 1, Spilosite, Spec. 32861 Lake Superior Division U. 8. Geol.
Survey, ‘from Mansfield, Michigan, Dr. H. N. Stokes, eae

No. 2, Spilosite, Spec. 32827 7 Lake Superior Division U.S. Geol.
Survey, ‘from Mansfield, Michigan, Dr. H. N. Stokes, analyst.

Desmosites—Under the desmosites are incinded contact
products composed of the same mineral constituents as the
spilosites, but which instead of being spotted show a distinctly
banded structure. That these are very closely related to the
spilosites, and that in fact they grade into each other is shown
by the study of one rock. A section from this, examined
under the microscope, shows irregularly rounded areas which
consist of ragged bunches of chlorite and aggregates of rutile
and epidote, with some flakes of biotite lying in a quartz-feld-
spar mass. As these spots increase in number they approach
each other and unite, forming streamers which in their turn
unite and form bands. No analysis has been obtained of the
desmosites, as it was so evident that they contain nothing dif-
ferent from the spilosites:

Adinoles.—Thus far chlorite has been the chief dark constitu-
ent of the contact products mentioned, whereas actinolite is

* C was not determined.

88 J. M. Clements—Study of Contact Metamorphism.

either totally absent or else is present merely as an accessory.
In the adinoles actinolite is the characteristic constituent.
The minor constituents, as a rule, are more uniformly distrib-
uted than in the spilosites. However some spots are present
and are composed essentially of actinolite. The actinolite is
present in sheaf-like growths, lying commonly in an exceedingly
fine-grained mass of quartz and albite, with some flakes of
chlorite and grains of epidote. The adinole is rendered rather
dark by minute black specks which are disseminated through
it. In places these are collected in irregular or lenticular heaps.
They seem to be carbonaceous matter.

ANALYSIS OF ADINOLE FROM MANSFIELD, MICHIGAN.
(By Mr. Geo. Steiger, U. 8. G. S.)

DIO eta 1 a a eee ee 74°16
A J mea as > ROS Mints yf ey 2 BST)
ANOn. Ss 11°85
Me Oi ee soe ee a eye 82
SEMEL) NaI AE Ni aii tan NES! Nl 1°66
MEO soe Ae ee ee 06
(GE © aan een) Ate atnc inter set 2°10
aie ae aie ee ae none
MeO e522 SAE eee ee ae 2200)
Re Oynxe ae cS.) SRR gs eee 15
Wean@ eco of a5. pn) Tage let gle ee 6°57
EH Ovat 100°-—,' 5 en ee 05
BsOxat 1005 4.4 oleh be ep ae "52
BO pee 12 365 cts ake a ae 08
COA Sono u 8 hs eee 09
Gig Fee eh he ee 18

TOtalcc 2 ee 2 eee 100°76

Other varieties of the contact rocks.—There is still another
kind of contact rock in which actinolite is the chief dark
constituent, and in this the actinolite is mainly collected in
bands. This rock thus corresponds to the desmosites (banded
chlorite rocks) in structure, though differing from them in
mineralogical composition.

The chlorite and actinolite contact rocks may be expected
to grade into each other, and such a gradation is shown in one
specimen, in which actinolite and chlorite are present in about
equal quantity. The actinolite occurs in crystals and sheaves,
forming spots; whereas the main mass of the thin section sur-
rounding the spots is formed by chlorite as the dark silicate,
associated with feldspar, quartz, and some epidote.

Comparison of analyses—F rom previous determinations in
other regions it is well known that the adinoles are next to the

J. M. Clements—Study of Contact Metumorphism. 89

contact, while the spilosites (and desmosites) are intermediate
between them and the clay-slates. The following series of
analyses are arranged in the order of approach to the dolerite,
as determined by the character of the rocks:

1 2. oF 4
si0, ee Se see Ne 1 60°28 O22" ek 74°16
Mey ii Oh 69 1°70 92 37
MORO Roe hie oe 22°61 19°00 19°35 11°85
le eee a none none
NO eerie 2 2°53 3°31 1:29 82
one Ee "45 7°19 Beal 1°66
oC ea ea a aa trace trace trace 06
NO) en eee ea ais 155 171 2°10
ee rey tee "04 trace none none
SS Se ae trace trace
iC) en 3°29 4°35 2°10
Os os IS ee ae hei “70 22 15
Na,O 46 Se Oe ae ee *b4 6°72 8°22 6°57
me eet trace none
H,O at 100° —_..- ‘60 aA. 18 °05
MONA 100° 4 2. oe. 3°62 43:26 42°34 "52
PAO eee 52 "08 15 04 08
Ua oe eee none none none 09
BiamdisO sl none none

ALS Se ee ‘97 18
Ce Ee ae ee none none
ee trace none
Ui ee 99 57 99°72 99°76 100°76

No. 1, Clay-slate, Spec. No. 32497 L.S. D., U.S. G. 8. Analyst
Mr. Geo. Steiger.

No. 2, Spilosite, Spec. No. 32861 L. 8. D., U. 8. G. S.- Analyst
Dr. H. N. Stokes.

No. 3, Spilosite, Spec. No. 32827 L.8. D.,U.8.G.S. Analyst
Dr. H. N. Stokes.

No. 4, Adinole, Spec. No. 32465 L. 8. D., U.S. G. 8S. Analyst
Mr. Geo. Steiger.

In these analyses the usual increase of silica as the dolerite is
approached is at once noticeable, and hand in hand with it
goes the diminution in percentage of alumina and iron oxides.
The content of water and carbonaceoust matter also suffers
diminution, as was to be expected.

The most noteworthy difference between the clay-slate and
the contact rocks is shown in the relations of potash and soda.
This is well brought out in an examination of analyses Nos. 1

* 20 at 110°. + H20 above 110°.
¢ The C in Nos. 2 and 3 was not determined.

90 SM. Clements—Study of Contact Metamorphism.
*

and 2. It will be seen that there is only about 4th as much
potash in the contact rocks as in the normal clay slate; while,
on the contrary, about 12 times as much soda as there was in
the slate has been added to the contact rock. This causes a
reversal of the relations of the soda and potash, so that, whereas
in the clay-slate there is present 10 times as much potash as
soda, we find in the contact rock taken as an example very
nearly 10 times as much soda as potash.

Comparison of Mineralogical Composition.

It will be sufficient for our present purpose to consider the
clay-slate, on the one hand, and the spilosite and adinole on
the other. To recapitulate briefly, in the clay-slate we have
quartz (clastic), white mica, actinolite, rutile, and an indeter-
minate interstitial material of finely granular character, discol-
ored by carbonaceous and ferruginous matter. This intersti-
tial material we may consider as that remnant of the original
feldspar, quartz, biotite, and iron ore dust in the original clay
which was not used in the production of the white mica, actino-
lite, and rutile now found in the clay-slate. Original chemical
differences in the slate are shown by the arrangement of the
various materials in bands, certain bands containing a greater
quantity of white mica; hence, the conclusion that these were
probably richer in alkali than the adjoining bands, which are
poor in white mica.

In the spilosite and adinole we find present quartz—but not
clastic—white mica, actinolite, and rutile, as in the clay-slate.
But in addition there is albite, chlorite, epidote, and biotite.
There is nothing in these rocks which shows elastic origin.
The original sediment has been completely recrystallized and
there remains only the banding to point to the sedimentary
origin of the rock. The clastic quartz has been destroyed by
solution or aqueous fusion, and has been recrystallized, and —
now forms a mosaic with the albite. Albite, for whose pro-
duction some of the silica of the clastic quartz may have been
used, is present in the rock in considerable quantity, in contra-
distinction to its total absence from the slate. The initial
quantitative chemical differences existing in the clastic rock are
also shown in the recrystallized rock by differences in the min-
eralogical composition,a large quantity of chlorite, for instance,
forming a band adjacent to a band poor in chlorite, thus indi-
cating a richness of the first band in magnesium. The great-
est difference existing between the practically unmetamor-
phosed slate and the contact product is the evidence in this last
of complete recrystallization, and the presence of albite in large
quantity. |

J. M. Clements—Study of Contact Metamorphism. 91

Cause of Chemical and Mineralogical Changes in the Rocks
described. |

The changes which have taken place in sedimentary rocks as
the result of contact with igneous rocks have given rise to a
great deal of discussion. The one school, notably represented
by the French geologists, contends that there is an actual
transfer of material from the igneous rock to the rock intruded
with more or less complete resorption of the latter; whereas
the other schoo] has maintained that the indisputable meta-
morphism observed in such cases is due chiefly to the action of
the so-called mineralizers. Studies upon a series of contact
rocks similar to those here presented have given foundation for
the view upheld by Roth,* Zirkel,t and others, that in the case of
contact metamorphism produced by basic rocks an actual trans-
ference takes place. This view has been recently very strongly
advocated by W. Maynard Hutchings,t who has described
some interesting products, which result from the contact of the
Whin Sill, which still further support it.

The very considerable changes which are shown by the
above analyses to have taken place in the metamorphism of the
slates, especially the change in the amount of silica and
soda resulting in the production of albite in large quantity,
seem to add weight to the supposition that in such contacts
an actual transfer of material, possibly in the form, as has
been suggested by others, of a soda silicate, does take place
from the basic igneous rock to the intruded slate.

Manpison, WIs.

* Chemische Geologie, by J. Roth: Berlin, 1890, vol. iii, p. 145.

+ Lehrbuch der Petrographie, by F. Zirkel: Leipzig, 2d edit , vol. ii, 1894, p. 722.

t Notes on the composition of clay-slates, ete.. and on some points in their
contact metamorphism, by W. M. Hutchings: Geol. Magazine, vol. i, dec. 4, 1894,
p: 15:

An interesting contact rock, with notes on contact metamorphism, by W. M.
Hutchings: Geol. Mag, vol. ii, dec. 4, 1895, p. 122-131, 163-169.

92 H. F. Osborn—Origin of Mammals.

Art. XI.— The Origin of Mammals ;* by Henry F.
OSBORN.

SINCE the source of the Mammalian phylum was later than
that of the Reptilian phylum in the Permian and possibly Car-
boniferous, we are sure that it extended at least very far back
into the Triassic, and the Triassic was apparently a time of
great continental connections and of consequent wide geo-
graphical distribution even of land types.

In presenting a phylogenetic chart which traces the ancestry
of Mammals to the Upper Permian, I am quite as conscious as -
the most conservative zoologist present that we are not on
sure ground, that this is a castle of cards liable to fall at any
moment. Yet such a chart, volving as it does numberless
hypotheses, is necessary to set forth certain ideas.

Clearing the way for this discussion, this chart first replaces
the general view defended by Huxley in 1880, of a genetic
succession between three sub-classes of mammals, which has
become a matter of creed with many zoologists. In the last
eighteen years nota scintilla of evidence has arisen to show
that Placentals are descended from Marsupials, and of late,
evidence has been coming in directly against this view. In
fact zoo-paleontology now indicates that Marsupials and
Placentals are parallel phyla, arising from a common stock,
while Monotremes are so different that they may even be con-
sidered diphyletic, or derived independently from the Reptilia,
as maintained by Mivart among others, and latterly by Seeley.

So far as major classification is affected by this conception,
it appears therefore that we must revert to Gill’s divisions of
1872, constituting only two sub-classes of Mammals, namely:

A—EHutheria | and B—Prototheria.
Marsupials, Pla- Monotremes.
centals.

Marsupials are less primitive than the placental Insectivores,
and Marsupials and Placentals are certainly far nearer each
other than either are to the Monotremes.

To guide our speculation in the unknown pre-Tertiary
period, we may gather certain positive principles from the
known evolution of the Tertiary mammalia. First, we know
that adaptive radiation, characteristic of all vertebrates, and
beautifully illustrated among Reptilia, is in a very high degree
distinctive of Mammalia, because of their superior plasticity.t+

*Opening the discussion before the International Congress of Zoologists at
Cambridge. (In conjunction with Professor H. G Seeley.)

+ We know nothing of Africa, whether it enjoyed a radiation of its own or bor-
rowed its fauna from other continents.

H.. F’. Oshorn— Origin of Mammals. 93

There is the (1) Marsupial radiation of Australia (Meteu-
theria)*—now passing its prime, then the (II) Zertzary Placen-
tal radiation of the Northern Hemisphere (Ceneutheria), and
another quite independent (III) Tertiary placental radiation
in South America, (rendered less pure in course of the Ter-
tiary period by migration from I and II). IV, There is the
entirely distinct and archeic Cretaceous Placental Radiation
of the Northern Hemisphere (Meseutheria), which extends
into the Tertiary, and may have given origin to II and III,
although as yet we have no direct proof of it.

We mark the fact that the above radiations are all of ordinal
rank, for the Marsupial radii, although termed families, are
adaptively equivalent to several Placental orders.

If we apply these same principles to the Jurassic, we appar-
ently have evidence of a more fundamental (V), Swb-class
radiation of Placentals and Marsupials, and probably Mono-
tremes, of world-wide distribution. (This involves a controverted
question, to which [ shall revert, for by some zoologists these
animals of the Purbeck Clays, Como Beds, and Stonesfield
Slates are considered exclusively Marsupials and Monotremes.)
Finally far back in the Perm-Trias we certainly observe (VI),
the Theromorph or Theriodont Reptilian radiation, spurs of
which may have given rise to the Mammalia.

The focal-types, or most primitive forms of the radiations,
I—LV, were certainly small, terrestrial, clawed, wnsectivorous
or omnivorous forms. It is noteworthy that in the evolution
of each radiation, so far as we know at present, land types and
organs are invariably primitive, and water types and organs are
secondary, exactly as we find it among the Reptilia. In fact
we have not found a single instance in which a mammal or
reptile series is known to be transforming from a water into a
land type; it is always the reverse. There is certainly no evi-
dence for a cetoid (Albrecht) stem of the Mammals. Again it
is obvious that neither carnivorous nor herbivorous types with
highly specialized or reduced teeth and feet can be so central
as insectivorous and omnivorous types. In fact the Insecti-
vores among Placentals, and Opossums among Marsupials, are
the only animals which have preserved the dental prototype
close to that of the Promammal.

The backward convergence of all Tertiary mammalia to a
Creodont stem, as indicated in the chart, affords the clearest
demonstration of the existence of an earlier insectivorous stem,
and it not only accords with the above principles, but we are
thus carried a step further to realize that the Creodont ancestor
must have been a generalized Insectivore.

*The Eutheria may embrace the Meteutheria or Marsupials, the Meseutheria
or primitive Mesozoic Placentals, the Cenoutheria or Tertiary Placentals,

94 H. F’. Osborn— Origin of Mammals.

To reconstruct the ancestral Eutherian skeleton, therefore,
we should divest the Creodonta and Insectivora, in their
anatomy and development, of all their secondary specializa-
tions, and combine all their truly primitive characters. This
ideal Eutherian is nearer the pro-Marsupial than we supposed ;
it has the following chief characters: Head large, body rela-
tively smal]. Anterior nares terminal, face elongate. Insectiv-
orous or omnivorous habit. Molar teeth pointed, tritubercular.
Typical dental succession. Vertebree with intercentra. Dorso-
lumbars not exceeding 20. Back arched, tail long and powerful.
Scapula and ilium narrow, acuminate. Humerus with powerful

HYPOTHETICAL PHYLOCENY oF THEMAMMALJA
AS DERIVED FROM PAL ONTOLOGY
RECENT |INSECTIVORA NOTREMATA

JURAS. INSEC are TRICONODONTA

a PLACENTALIA------AARSUP, ALL

CARNIVOROUS) yl
1,DICYRODOHTIA 2,THERIODONTIA 3, abe ane a ,COMPHODONTIA
=

PROCANOSAURIA= _ (THEROMORA=
RHYNCHOCEPHALIA ( LaNenie tA 1A
(MONOCONDYLIA) (MOWO. SDICONDT DICONDYLIA)
A {GOTYLOSAURIA- AMPHIBI,

PARIASAURI
kata ty Be (MONOCONDYL A (o1conDy La)

i y Veal baa
—--=-<---

\ gepritia ) BATRACHIA

STEGOCEPHALA= LABYRINTH
(DICONDYLIA)

deltoid crest, condylar crests and entepicondylar foramen.
Femur with three trochanters. Feet plantigrade. Centralia
and tibiale. JF ore-limb more or less prehensile, elbows everted.

Now the Creodonta are members of that Cretaceous radia-
tion, IV, which we find in full bloom in the Basal Eocene,
mingled with the dying group of Multituberculates, which
belong to the Jurassic radiation, V.

In the Mid—and Uppermost Jurassic (Stonesfield, Purbeck
and Como Beds) however, occur all three types, which theoreti-
cally constitute radiation V: first, the Z7rzconodonts primitively
but typically Marsupzal in structure ; second, the Jnsectwora
primitiva, Seaacelsy Placental in type, insectivorons and
without a single Marsupial character either in jaw or teeth;
third, the Multituberculata, whose sub-class position is now
assumed to be Monotreme.

H, F. Osborn—Origin of Mammals. 95

Always keeping in mind that our direct evidence here is of
the most limited character, since we have neither skulls nor
skeletons, only teeth and jaws, we are tempted to hypothetically
connect the Creodonta with the Insectivora Primitiva, and to
assume that there existed in the Mid-Jurassic, as above stated,
the well advanced radiation V of the two sub-classes of HUTHERIA
and ProTorHERIA as defined by Gill.

The problem of the Origin of the Mammals now resolves
itself into the connections between these two sub-classes either
with the Reptilia or Amphibia, and we turn back to the three
contemporary Upper Permian reptile groups:

1. Pareiasauria or Cotylosauria, \jand animals with a solid
skull and replete with Stegocephalian or Amphibian characters,
certainly the most primitive reptiles.

29. Proterosauria or Proganosauria, with an open two-arched
skull, specialized reptiles, which have apparent affinities with
the Crocodilia, Dinosauria, Rlyncocephalia and Squamata (Doli-
chosauria, Mosasauria, Lacertilia and Ophidia).

3. Theriodontia or Theromora (Dicynodontia, Cynodontia and
Gomphodontia), with an open, single-arched, skull (as in the
Chelonia, Plesiosauria, Icthyosauria and Mammalia).

The Theriodontia as perceived by Owen in 1876, and now
fully confirmed by Seeley, are astoundingly mammalian in
type. They are essentially quadrupedal, long-limbed, terres-
trial reptiles, totally dissimilar from all other reptiles, and
standing far away from them. They also present an advanced
stage of functional radiation in tooth and skull structure in
adaptation to carnivorous, omnivorous and herbivorous habits,
to which, when known, their skeletons will probably be found
to conform. The shoulder girdle is of Monotreme type; in
other features of the skeleton they are strikingly like the
ancestral Eutherian, described above. The skull may in fact
be reduced to the Eutherian type by the coalescence of the
prefrontals, postorbito-frontals and quadrates with the adjacent
elements, and by the loss of the ectopterygoids or transverse
bones. The teeth are promammalian in formula, and pro-tri-
tubercular or multitubereular in form. The skeleton so far as
known is partly Eutherian, partly Monotreme or Prototherian.

Two alternatives at this point present themselves: First: Are
the Monotremes directly derived from such a type as these
Theriodonts, leaving an independent derivation for the Euther-
jan or Marsupio-placental stock, as has been suggested by
Seeley? Second: Must we set aside the Theriodonts for an
Amphibian stem form on account of the numerous Amphibian
resemblances which Hubrecht and others find in the develop-
ment and anatomy of the lower Mammalia ?

96 Hi. EF. Osborn— Origin of Mammals.

The first question is now unanswerable, although at present
the evidence that the Mammalia are diphyletic is certainly
considerable. Second, the investigation of the placental
ovum rises to extreme importance; if the Eutherian ovum is of
Amphibian type, the Mammalia are certainly diphyletic, for
the Monotreme ovum is certainly of reptilian type. This
point may be met by supposing that at the time the
Marsupio-placentals were given off, the Theriodontia conserved
a number of Amphibian characters, which were at the time or
subsequently lost.

The problem of most immediate concern, therefore, is
whether the Theriodontia are actually the long-sought Pro-
mammalia of Heckel, Hypotheria of Huxley or Sauromam-
malia of Baur, or whether they present afresh instance of
extensive parallelism due to the assumption of habits analogous
to those of Mammalia. Professor Seeley has just presented
the latter view,* and it is certainly true that none of the
known Theriodontia fill the characters outlined above as those
we must look for in the Promammal or Eutherian stem, for
they are all too large and too specialized.

There are however grounds for the more sanguine former
view, that the Theriodontia are the Hypotheria or Promam-
malia, because it appears that wzthin the order may well have
existed some small insectivorous types, far less specialized in
tooth structure than either the carnivorous Cynodonts or her-
bwvorous Gomphodonts, as one of those conservative spurs of
adaptive radiation which form the focus of a new progressive
ty pe.
The problem will therefore be settled by additional discovery
and knowledge of the structure of the diverse types which
undoubtedly composed this remarkable group.

Notr. Wecan happily preserve that part of Huxley’s speculation upon the
origin of the Mammals which pictured the Insectivora as nearest the ancestral
type. The Amphibian (or Stegocephalian) origin of the Mammalia which he
defended, is wholly set aside, if the view here taken is correct, for the Theriodontia
are certainly not Amphibia in any sense.

*“ Anomodonts are not the parents of Mammals, but a collateral and closely
related group. The common parent of both may be sought in rocks older than
Permian, perhaps in Silurian or Devonian strata.” Printed abstract of Professor
Seeley’s argument, p. 3.

Penfield and Foote—Composition of Tourmaline. 97

ArT. XII.—On the Chemical Composition of Tourmaline;
by 8. L. PENFIELD and H. W. Foote.

INTRODUCTION AND HistToricaAu.—There is probably no
common mineral whose chemical composition has proved more
perplexing and been so little understood as tourmaline. Some
reasons for this are, first, that the mineral presents certain
peculiarities in chemical composition of an unusual nature;
second, the analysis of tourmaline has been one of the difficult
problems of analytical chemistry, hence reliable data for the
-ealeulation of the formula have not been easily obtained; and,
lastly, although good analyses have been made, the results have
not been thoroughly relied upon, nor have they been inter-
preted to the best advantage. The present investigation was
undertaken, therefore, with the hope that by making a few
analyses with the utmost possible care on tourmalines of excep-
tional purity, it would be possible to find a satisfactory explana-
tion of the chemical composition of this interesting mineral.

In order to appreciate the problem in hand, it will be neces-
sary to review briefly the work and the results of previous
investigators.

The analyses of Vauquelin and Klaproth, made in the early
part of this century, were naturally defective, because at the
time they were made, lithium was unknown, it had not been
discovered that tourmaline contained boron, and analytical
methods were not perfected.

In 1818 the presence of beron was detected by Lampardius,*
and in the same year Arfvedsont discovered the new alkali
metal lithium, and showed its presence in spodumene, petalite
and tourmaline.

In 1827 Gmelint published analyses of ten varieties of tour-
maline, but his results led to no satisfactory formula, although
the essential constituents of the mineral, with the exception of
the boric oxide, were determined with a considerable degree
of accuracy.

In 1845 Hermann§ published the results of four analyses of
tourmaline from Russian localities. He proved conclusively
that the iron was ferrous, and not ferric as considered by pre-
vious investigators. Boric oxide was not directly determined,
but estimated by difference, and the results compare favorably
with the direct determinations made by our present methods.

* Ann. d. Phys. u. Chem., xxx, p. 107.
+ Schweiger’s Jour. d. Chem. u. Phys., xxii, p. 111.

¢ Pogg. Ann., ix, p. 127.
§ Journal fiir prakt. Chem., xxxv, p. 232.

Am. Jour. Sc1.—FourtH Series, Vou. VII, No. 38.—FEBRUARY, 1899.
7

98 Penfield and Foote-—-Composition of Tourmaline.

He was the first to point out that silica and boric oxide are
present in the definite molecular proportion 4:1. He errone-
ously decided that tourmaline contained carbon dioxide, the
reasons being as follows: It was generally believed at that
time that the mineral contained no water, as stated by Hermann
“die Turmaline keine Spur davon enthalten,” and it is true
that when fragments are tested by the usual method of heating
to redness in a closed tube no water is obtained. It is only
when the material is heated zntensely, best as fine powder, that
hydroxy] is decomposed and water given off. When Hermann
dissolved fragments in a borax bead he observed that a gas was
evolved, and, since it was believed that this could not be water
vapor, he supposed that it must be carbon dioxide. Pains
were taken to fuse some of the mineral in a tube with borax,
and to conduct the gas into lime water, by which treatment a
precipitate was obtained which effervesced with acids, but it is
safe to assume that the carbon dioxide thus detected was de-
rived from improperly purified air and not from the mineral.

In 1850 Rammelsberg* published the results of the analyses
of thirty varieties of tourmaline. The execution of such a
large number of analyses must be regarded as a very great
undertaking, since at that time gas and many facilities of our
modern laboratories were not available, and many methods of
analysis were not perfected. Special evidence is given that
great care was taken in the selection of material for analysis
and in the analytical methods, which appear to have been well
chosen and reliable in character. The analyses, however, were
defective in several important particulars. Thus the iron was
regarded chiefly as ferric. Believing, like Hermann, that
tourmaline contained no water, and having detected the pres-
ence of fluorine in some varieties of the mineral, he supposed ~
that the considerable loss on ignition which occurred was due
to the volatilization of silicon fluoride, and from this loss he
estimated fluorine to be present in amounts varying from 1°30
to 2°51 per cent. Direct determinations of boric oxide were
made in three cases only, and in the remaining analyses this
important constituent was estimated by difference. Although
the analyses led to no satisfactory formula, they indicated cer-
tain prominent characteristics of the mineral, namely, the great
variation in the relative amounts of aluminium, iron, magne-
sium and alkalies, and the nearly uniform amounts of silica
and boric oxide.

Fully realizing certain defects in his earlier analyses, Ram-
melsbergt published in 1870 a revision of his former paper.

* Ann. der Phys. u. Chem., lxxx, p. 449 and Ixxxi, p. 1.
+ Ann. der Phys. u. Chem., cexv, pp. 379 and 547.

Penfield and Foote—Composition of Tourmaline. 99

At this time it was shown that all varieties of tourmaline con-
tained chemically combined water, and the amount of water was
estimated from the earlier determinations of the loss on igni-
tion after making certain corrections for volatilization of sili-
con fluoride. He found that the iron was chiefly if not wholly
ferrous, and recalculated accordingly his earlier results. Six
direct determinations of boric oxide were made, and it is
pointed out that these amounts correspond closely with the
indirect determinations by difference. As a result of the
revision, Rammelsberg reached the conclusion that all tourma-
lines are derived from the acid H,SiO,. In this he considered
the hydrogen atoms to be replaced by metals of different val-
ences, or, in other words, he regarded tourmaline as composed
of a mixture of the following molecules:

R’ SiO, Re = Na; Ky las and.

Rh’ Sid, R” = Fe, Mg, Mn and Ca.

i heey Oe B= Al ands:
Furthermore he decided that certain varieties correspond
closely to the special formulas

R’,Al,BSi,O,, :

4 nee. 4 11. |

4°20

B’,Al,,B,Si,O

R",Al,,B,Si,O,,

while he regarded others as mixtures of these two molecules.
By substituting hydrogen atoms for the metals and boron of
these special formulas, the acids become respectively H,,Si,0,,
and H,,Si,O,,, both of which are multiples of H,SiO,. He
concluded that the SiO, and B,O, are not present in a definite
molecular proportion, but that boron plays the part of a metal
and is isomorphous with aluminium.

In 1888 Riggs* published the results of twenty analyses of
various types of tourmaline from American localities. The
analyses were executed in the laboratory of the U. 8. Geologi-
cal Survey at Washington, and bear every evidence of being
made with the precision and care characteristic of the analyti-
eal work of that laboratory. Boric oxide, water and ferrous
oxide were determined directly by reliable methods, and a high
degree of accuracy is claimed for the analyses. A careful
description of the quality of the material analyzed is not
given, and although it is to be supposed that great care was
taken in its selection, the following statement has left this in
doubt: ‘The analyses do not represent ideal compounds, but
are made of material more or less impure....” Riggs
concludes that the analyses give “as a general tourmaline
formula the simple boro-orthosilicate R,BO,2Si0O,” which is
expressed graphically as follows:

* This Journal, III, xxxv, p. 35.

100 Penfield and Foote—Composition of Tourmaline.

/ BO,
R, = SiO,

R, Z si0 DF.
He supposes R to include H, Li, Na, K, Ca, Mg, Fe, Al and
small amounts of the Al=O or possibly Al-OH radicals. It
may here be stated that the foregoing formula is identical in
type with the special formula R’,A],BSi,O,, (H,BSi,O,,) of
Rammelsberg, and considering boron as replacing hydrogen
like a metal the silicic acid from which the formula is derived
becomes H,,8i,0,, or H,Si0,. Riggs further states that, owing
to slight variations, the ratios give nearly the “equally simple
general formula R,,BO,28i0,,” stating that “ between these
two views there are at present no means at hand of deciding.”
It would seem, however, that the last formula is impossible,
for, considering hydrogen atoms as replacing R,,, the acid can
not be split up like other oxygen acids into silicic and boraciec
anhydrides and water. There are also given the following
special formulas for three pronounced types of tourmaline:

I. Lithiatour. 12Si0,,3B,0,, 4H,O, 8Al,0,, 2(NaLi),O.
If. Iron tour. 128i10,, 3B,0,, 4H,0, 7Al],0,, 4FeO, Na,O.
Ill. Mg tour. 12S8i0,, 3B,0,, 4H,O, 5A1,0,, 28Mg0O, #Na,O.
By substituting hydrogen atoms for the metals in these spe-
cial formulas we obtain:

I. and II. H,,B,Si,,0,, or H,,B,Si,O

PSA ON

IIL. H,,B,Si,,0,, of H,,,B,81,0,,9.

58ye0 6 12 62

Soon after the appearance of Riggs’ article, Wilfing* recal-
culated the results of these twenty analyses and concluded that
all tourmalines may be regarded as isomorphous mixtures of
two aluminium silicates, “Alwmosdlicate,” of the following
composition.

I, Alkali tourmaline 12Si0,. 3B,0,.8Al,O,.2Na,O . 4H0,
II. Magnesia tourmaline 12810,.3B,O0,.5A],0,.12Mg0O. 3H,0.

In these formulas it is assumed that the isomorphous ele-
ments K and Li replace the Na; Fe’” the Al; and Fe’’, Mn
and Ca the Mg. By substituting hydrogen atoms for the
metals in the foregoing formulas it is found that they both are
derivatives of the same acid, H,,B,Si,,O,, or H,,B,Si,O,,. The
conclusions derived by Wiilfing are that, although in most
cases the results of the analyses agree with the percentage
values calculated from his formulas, the agreement is not
always satisfactory. This he ascribes to the possible need of a

* Mineralogische und petrographische Mittheilungen, x, p. 161.

Penfield and Foote—Composition of Tourmaline. 101

third formula; to possible inaccuracies in the difficult ferrous
iron determinations; and in part to the fact, as stated by Riggs,
that “the analyses do not represent ideal compounds... .”
He therefore considers it necessary that further analyses of
more carefully selected material should be made.

At about this time also Scharizer* published analyses of three
varieties of tourmaline from Schiittenhofen, Bohemia, and dis-
cussed his results in connection with the analyses of Riggs.
His final conclusion is that, with the exception of the green
varieties, tourmalines possess a chemical constitution which can
be expressed by the general formula

Be y YS TLE O
ek || RE, | aysion,| co, HO, ry, |.

This formula is certainly difficult to comprehend, and, if we
understand it correctly, the univalent radicals BO and HO are
isomorphous with fluorine, and these constituents, taken twice,
can replace oxygen.

In 1889 Jannasch and Kalbt published the results of nine
tourmaline analyses, in which the water and boron were deter-
mined directly. These investigators derived from their anal-
yses the general formula R,. BO,.(S8iO,),, for which the fol-
lowing structural formula was proposed :

Om

7S jae 2

This formula, R,.BO,.(SiO,),, is identical with the one pro-
posed by Riggs, and essentially like the special formula
R’,A1,BSi,0,, of Rammelsberg. The following special formu-
las are also given:

I. Li tour. 24Si0,. 6B,
II. Fe tour. 24810, .6B,O
Il. Fe-Mg tour. 24Si0,.6B,O

2

O,.15AI1,0,.4FeO . 4(Li,O, Na,
.14A1,0,.9FeO. 2Na,0. 7H,

.13A1,0,12MgO . 2Na,O. 7H,O

These special formulas in their general type are similar to
those proposed by Riggs. By substituting hydrogen atoms
for the metals they all reduce to one and the same acid,
Pa. D,291,,0,., or Fileore)...

* Zeitschr. fiir Kryst., xv, p. 343.

+ Berichte der deutschen chemischen Gesellschaft, vol, xxii, p. 216. Also In-
augural] Dissertation, Geo. W. Kalb, Gottingen.

): 78.0

2

3
3
3

102 Penjield and Foote—Composition of Tourmaline.

There is evidence that Jannasch does not place great confi-
dence in the foregoing formulas, for in Hintze’s Mineralogy*
the.composition is expressed as follows:

I. Lithia tourmaline _§i,,0,,B,AJ,,(Na, Li),H,
II. Iron tourmaline 8i,,0,,4B,Al,,Fe,Na,H,

III. Magnesia tourmaline Si,,0,,B,Al .Mg,,Na,H,

12 69

These formulas are essentially different from the ones first pro-
posed, which may readily be seen by substituting hydrogen
atoms for the metals and comparing the resulting acids, as
follows: 7

I. H,,B,Si,,0,, or H,,,B,Si,0

12 63 4 ~ 21

Il. H,,B,Si,,0,., or H,,B,Si,O,,3

12 67

TlH BS0. or Hg Bis OF

Soon after the appearance of the articles of Riggs and of
Jannasch and Kalb, the results of their analyses were recalcu-
lated by Goldschmidt,+ and the conclusion was reached that
tourmaline may be regarded as containing the two following
molecules:

I. Alkali tourmaline R’,,R'" 51
II. Magnesia tourmaline R’,,R”,,R’

oi Our

81,,0

I ]
44 31 162

These formulas are both derived from an acid H,,,8i,,O,,, or
Oa:

In 1893 Rheineck,t after calculating the results of the many
tourmaline analyses, concluded that the composition of all
tourmalines may be expressed by formulas of the following
type, in which Al, appears as a constant:

I. Alkali tourmaline Alsi, BEG;
III. Magnesia tourmaline Al,Si.B,M,H,O,,, ete.

4~ 259

M represents the bivalent metals Fe, Mn, Mg and Ca. H
includes Na, Li and K. Among the numerous examples he
stated that the composition of the black tourmaline from
Pierrepont, N. Y., may be represented by either of the follow-
ing expressions : ;
{114 A1,Si,B,M,H,0,, 10 Al,Si,B,H,0,,
j 10 Al.Si,B.O.. or 193 AlSi.B,M.H,O,,
21 Al,Si,B,M,O,,
* Vol. ii, p. 311 (communication from manuscript by Jannasch)..

+ Zeitschr. fir Kryst, xvii, pp. 52 and 61.
t Zeitschr. fur Kryst., xxii, p. 52.

Penfield and Foote—Composition of Tourmaline. 103

By substituting hydrogen atoms for the metals in these ex-
pressions, and multiplying by the factors, 114 and 10 in the
one case, and 10, 93 and 21 in the other, the acids become
Se Oe and Ee Bo O14. Oe. or, when simplified,

po 191,0,,., and H,,,B, ,Si,O,,,- The improbable nature of
such acids is evident. In the opening paragraph of his article
Rheineck stated that the obscure and complex chemical rela-
tions of this mineral have necessitated a series of speculations
and calculations extending somewhat interruptedly over a
period of many years, in order to arrive at results such as those
embodied in the foregoing formulas.

In 1895 Clarke* discussed the constitution of tourmaline and
proposed the following formulas:

iL. 2.
yw > jh’, / 10,= Mell
Al—Si0 = Al Al—siO = MgH
\$10,= Al— BO, \810,= Al— BO,
| |
Al—BO,= NaH Al—BO,=NaH
| |
/310,=Al—BO, / Si0,=Al—BO,
Al—Si0,= Al Al—Si0 = Al
Posi) Al SSi0— Al
a 4,
Poi — Met / 10 = Mg
Al—Si0 = MgH Al—Si0 = MgH
\810,=AlI—BO, \S10,=Al—BO,
Al—BO,=NaH — Al—BO,=NaH
| |
/ si0,=Al— BO, / si0,=Al—BO,
Al—Si0,= MgH Al—Si0 = Mg
X10 = Al S10 — Molt

Clarke assumes variations from these formulas in that Fe’’’
and Cr can replace the Al; Fe” and Mn the Mg; Ca the
NaH; and small amounts of F the BO,. Proof of a consti-
tution corresponding to the formulas cannot of course be
expected, but it is doubtful whether aluminium could exercise
such varied functions as the formulas indicate. The grounds
for believing that fluorine can replace BO, are not stated. The
acid from which all these formulas are derived is H,,B,Si,O,,
or H, ,B,8i,0,,¢.

Lastly Grotht+ has adopted the formula of Jannasch
[Si0,],. BO,. R’,, but interprets it as follows:

* Bulletin of the U. 8. Geological Survey, No. 125, p. 56.
+ Tabellarische Uebersicht der Mineralien, 4te Auflage, 1898, p. 117.

104 Penfield and Foote—Composition of Tourmaline.

=O
or [Si0,],[AlO.. BO]([Al0],, Mg, Fe, Na,, Li,, H,),

That the univalent, radical [AlO] can replace R’ does not
appear in the original article of Jannasch, and it makes a
decided difference whether three R’s are replaced by one atom
of Al or by three [AIO] radicals. The latter assumption ©
implies a basic character which tourmaline does not possess.

We have thus reviewed the work already done in order to
show the difficulties which this problem has presented. It is,
however, interesting to note how closely different investigators
come to one type of acid from which all varieties of tourma-
line are derived. For the sake of comparison these acids have
been reduced to four silicon atoms, and are given below, both
as borosilicic acids and with the boron replaced by hydrogen.

R ilesNer H,,B,Si,0,, or H, 51,0, =o so:
re “Ts HS Bs OF or EH 510F
(ad B.S, OF, or) Hysy@Oe
Rives | H,,B,8i,0,, or H,,5i,O,, (Irrational)
sé 4 Boe or. H, Si, 21
EH oO or He Oe
Jannasch and Kalb 1 1 Beto” es Be?
Wiilfing EH, BS OF or TSO
Goldschmidt i, STO

FB 0o.¢ OF Hy, 81,078. lmational)

Rheineck pe
yo, Lee or H,,. SuOes etc.

Clarke H,,..B,Si,0,,.. or H,.281,0

New ANALYSES..
_ Method of analysis.—The present investigation was under-
taken with the expectation that for the solution of the problem
in hand it would probably not be necessary to make a long
series of analyses but rather exceedingly accurate analyses of a
few carefully selected types of tourmaline. We therefore
prefaced our work by a careful study of those features of the
tourmaline analysis, which have proved most difficult. The
method proposed by Gooch* of distilling off the boron with
methy] alcohol and weighing it as calcium borate is very exact,

* Amer. Chem. Jour., ix, p. 23.

Penfield and Foote—Composition of Tourmaline. 105

but its application to an insoluble silicate, especially one con-
taining fluorine, needed careful study. Mixtures were accord-
ingly made of silicates to which weighed quantities of borax
and fluorite were added and the following conditions were
determined, which yielded the most accurate boron determina-
tions. The mineral was fused with from four to five parts of
sodium carbonate, the fusion extracted with water, and, with-
out filtering, an excess of ammonium carbonate was added.
The insoluble residue and the precipitate were filtered off, the
filtrate concentrated, acidified slightly with nitric acid and dis-
tilled with methyl alcohol. The residue from the sodium ear-
bonate fusion together with the precipitate produced by the
ammonium carbonate was fused again with sodium carbonate,
treated with water, filtered and distilled with methyl alcohol
after acidifying with nitric acid. About one half of one per
cent of boric oxide was obtained from the second treatment,
avd in no case did we succeed in obtaining an exact determi-
nation of the boric oxide without repeating the fusion. The
weighed mixture of calcium borate and oxide was in all cases
found to contain a small amount of fluorine. It was there-
fore dissolved in hydrochloric acid, and part of the lime
together with calcium fluoride was precipitated with sodium
carbonate. The precipitate was ignited, treated with acetic
acid and the resulting calcium fluoride weighed. The amount
of fluorine thus found never amounted to over 0-20 per cent.

Fluorine was determined by the modified Berzelius method
described by Pentield and Minor.*

Water was determined by fusing the mineral with sodium
carbonate in a combustion tube and collecting the water in a
weighed tube containing sulphuric acid, a method which has
been thoroughly tested and is known to give reliable results.

For the determination of the bases the mineral was decom-
posed by fusion with sodium carbonate and the silica separated
as usual. It was found by experiment that the amount of
silica volatilized by the small amount of fluorine in the min-
eral could practically be neglected. In one variety of tourma-
line two determinations of silica made by the Berzelius method
of fusing the mineral and a weighed amount of silica with
sodium carbonate, and separating the silica with ammonium
carbonate and an ammoniacal solution of zine oxide, gave 36°69
and 36°76 -per cent, while determinations by the ordinary
method gave 36°75 and 36°73 per cent. A similar conclusion,
that when the amount of fluorine is small it is not necessary to
separate the silica by the Berzelius method, was also reached
by Riggs.

* This Journal, III, xlvii. 1891, p. 387.
+ This Journal, III, xlviii, 1894, p. 31.

106 Penfield and Foote-—Composition of Tourmaline.

The filtrate from the silica was evaporated to dryness, mois-
tened with hydrochloric acid and repeatedly evaporated with
methyl alcohol to remove all possibilities of boric oxide being
precipitated with the bases and thus increasing their weight.

For the determination of the alkalies the Smith fusion was
employed and, after removal of ammonium salts, the residue
was treated with acid and methyl alcohol and evaporated to
remove any borate that might possibly be present. Lithium
was separated by the Gooch method of boiling with amyl
alcohol,* and was finally weighed as sulphate.

It was proved by careful qualitative tests} that the iron
was ferrous, and that, at the most, not more than traces of ferric
iron could be present. This statement holds good not only for
the varieties analyzed, but for all the varieties of black tourma-
line which were accessible to us.

Selection and preparation of material.—One of the varieties
selected for analysis was the white or colorless tourmaline from
De Kalb, St. Lawrence Co., New York. This was chosen
because according to the analysis of Riggs it represented
almost the extreme type of a magnesia tourmaline, and,
containing almost no iron, there could be no appreciable error
from a failure to estimate that constituent correctly. The
material was derived in part from a specimen in the Brush
Collection and in part from specimens collected by one of us
(Penfield) while connected with the U. 8. Geological Survey.
The clear, colorless, glassy material was most carefully selected
with the aid of a lens, ground and sifted to a uniform grain,
and suspended in methylene iodide. The specific gravity was
uniform, and the portion used for the analysis floated at 8-065
and sank at 8°033. As an additional precaution the grains
were treated with a mixture of hydrochloric and hydrofluoric
acids, which have almost no action even on finely pulverized
tourmaline, in order to remove any possible traces of adhering
calcite, tremolite or pyroxene, although these were not seen
nor believed to be present. It may be stated concerning the
final product that probably it was as pure as it is possible to
get a mineral substance.

Another variety selected for analysis was from the feldspar
quarries at Haddam Neck on the Connecticut River. Wonder-
ful tourmalines have recently been obtained from this locality,
and they are already well know to most collectors. . We are
indebted for our supply of material to Mr. Ernest Schernikow
of New York, who generously placed at our disposal an almost
unlimited supply of crystals of gem quality. We selected
small prisms, 2 to 4™™ in diameter, of a uniform rather pale
green color. They were of ideal purity, perfectly transparent

* Amer, Chem. Jour., ix, p. 33.
+ See note, p. 124.

Penfield and Foote—Composition of Tourmaline. 107

and without flaws. Any traces of foreign material that might
possibly be adhering to their outer surfaces were removed by
treating them for a considerable time with hydrofluoric acid.
The specific gravity, taken with the chemical balance, was
found to be 3-089.

We were prepared to extend our investigation by analyzing
other varieties, but having completed the above mentioned
two and finding that the results corresponded with those of other
investigators, 1t was decided that the data which would be
derived from further analyses would probably add very little
to the knowledge which we already possess.

The results of the analyses are as follows :

COLORLESS TOURMALINE, DE Kates, N. Y.

I. II. Average. Riggs.
SiO, ee 36°69 36°76 36°72+60= ‘612 4°00 36°88
TiO, Beet "06 "05 05 12
B O, pe. 10°86 10°77 10°31+70= "154 - 4701 10°58
Al,O, meee o-75 6 629°61 6 «29°68 17S] 8 741 ) 28°87
a "23 “A *22=—36= ‘006 54
MgO ——_ = 14:91 14:92 14:92+20= -°746 14°53
2 ile 3°47 3°50 3°49-—28= °125 $ 3:042 19°90 3°70
Lad rn 1:29 1:23 1:26+-31= -040 1-39
K,O 2 an 07 °03 °05+47= ‘002 =i
eee 23 Bee ULES) (2-98- 9= 331 3°56
Je ee 92 "95 “93+19= +049 ) "50
tO1-11 100°83
O equivalent to F__--_- “39 a
100°72 100°62

GREEN TOURMALINE, HADDAM NECK, CONN.
Brazil.
iJ Il. Average. Riggs.
S10, Bee. 36°87 37°05 3696-60 = 616 4°00 37°39

er. 08 03 ‘03 ?
B,O, ----11°09 10°92 i1:00+70

=

Al,O,.---39°53 39°59 39°56+17 =2°327 ) 39°65
ReO fo. 2 - 212 «62°15 0=—_- 2°14+-36° = -059 *2°42
MnO .--- 1°96 2°04 2°00+35'°5= :056 | 1°47
_ Sal 0 eed Bs 15 "15+20 = ‘008 none
maw)... - 1:32 1:25 1:°28+28 = °046 $3:078 19°98 -49
a.O. 2.2913 .. 2°06 2°10+31 = 7068 t2°67
Li,0 .--. 165 163 1°64+15 = ‘110 | 171
BO. < 3°14 3:06 310+ 9 = ‘344 3 63
ee ee LOD 1:17 , 1713-419) =, 060. J 32

101°09 100°04
O equivalent to F.-.- ‘48 13

100°61 99°91

* Includes 0°15 per cent. Fe.O;. + Includes 0°25 per cent. K,0O.

108 Penfield and Koote—Composition of Tourmaline.

On treating the analyses according to the customary method
of deriving a formula (dividing each constituent by its molec-
-ular weight and finding the ratio of the quotients) it was
found that although the SiO, and B,O, were present in the
proportion of 4:1, no definite relation could be detected between
the silica, the different kinds of oxides, and the water. It was
decided, therefore, to determine the relative number of hydro-
gen atoms equivalent to the metals and thus learn the acid
from which tourmaline is derived. This was readily accom-
plished by dividing the constituents by appropriate fractions of
their molecular weights; for example, since the aluminium
atoms in Al,O, replace six hydrogens, the quantity of A1l,O,
was divided by one-sixth of its molecular weight, the FeO by
one half of its molecular weight, etc. Since fluorine is
regarded as playing the same role as hydroxyl, its ratio was
added directly to that of the hydrogen. The result of this
treatment is very satisfactory. The first analysis gives the
ratio of SiO,: B,O,:H=4:1:19°90, and the second) 4:4502;
19°98. Both ratios approximate so closely to 4:1:20 that
there can be no reasonable doubt that the acid from which
these tourmalines are derived is H,,B,Si,O,,. This formula
may seem at first somewhat complex, but it is not especially
so for a boro-silicie acid. It can not be simplified by division,
and it is based upon the very best kind of evidence, namely,
the close approximation to rational numbers of the two ratios,
which are derived from widely separated types of tourmaline.

Before discussing the possible constitution of this acid, it will
be shown to what extent the analyses of other investigators
confirm the results obtained by us.

REVIEW oF RAMMELSBERG’S ANALYSES.—The analyses
appear with certain variations in his different publications,
and the results which will be discussed here are given in his
Mineralchemie, 1875. These are quoted in the sixth edition
of Dana’s Mineralogy, and where there is occasion to refer to
individual analyses they will be designated by the numbers
assigned to them in the latter work. The same plan also will
be adopted for the analyses of Riggs, Jannasch and Kalb and
others. Thirty-five analyses are given by Rammelsberg, of
which No. 84 is not complete.

Adopting the same method of calculation as applied to our
own analyses, and using one-fourth of the silica as unity, the
ratios of SiO,: B,O, were found to vary between 4:1°21 and
4:0°71, and the ratios of SiO, to the total hydrogen atoms
between 4:16°5 and 4:19°7. The result of grouping the ratios
of the hydrogen atoms is as foliows :

Limits of variation, 16°5—17°5 17°5—18°5 18°5—19'5 19°%5—19-7
Number of analyses, 8 12 12 2

Penfield and Foote—Composition of Tourmaline. 109

The ratios of SiO, to the total hydrogen atoms show a wide
variation, and the hydrogens for the most part fall far below
20, which is required by the new formula. Rammelsberg,
however, claims that boron plays the part of a metal, replacing
aluminium, and that all tourmalines are derivatives of an acid
H,Si0, or H,,8i,O,,.. Replacing, therefore, the boron by hydro-
gens (B,O, is equivalent to 3H,O), it is found that the ratios
of SiO, to the total hydrogens then vary from 4: 21°9 to 4: 25-1.
The result of grouping the ratios is as follows:

Limits of variation, 21°9—22°5 22°5—23°5 23°5—24°5 24:°5—25'1
Number of analyses, 3 10 17/ -

The ratio of SiO, to the total hydrogens demanded by
Rammelsberg’s formula is 4 : 24, and in only seventeen, or one-
_half, of the analyses are the ratios 24+ 0°5, while they are
reasonably sharp, 24 + 0°2, in only seven cases. It would seem
therefore that owing to certain defects the analyses fail to give
decisive results. The ratios do not satisfactorily support
Rammelsberg’s formula nor the one which we have proposed.
In his latest publications* Rammelsberg claims great accuracy
both for his analyses and for the analytical methods which he
employed, and in justice to him it does not seem right to
ignore his results without endeavoring to discover the possible
defects of the analyses and the influences exerted by them
upon the ratios. In the first place, it is believed, for the fol-
lowing reasons, that his silica determinations are too high: it is
presumable that facilities were not at hand for producing the
intense heat necessary for expelling the last traces of water
from the silica; there is no statement that the silica was tested
as to its purity by volatilization with hydrofluoric acid; and,
lastly, a little silica might have been derived from glass beakers
and even from the distilled water, provided it was kept in glass
bottles or carboys. It is certain that the water determinations
are too low, and this is admitted by Rammelsberg. It is well
known that when a silicate containing fluorine and hydroxy] is
heated some hydrofluoric acid is liberated, and if the experi-
ment is tried in a closed tube the glass is etched, while proba-
bly the mineral also is attacked with the formation of silicon
fluoride. Rammelsberg,+ however, points out that only a por-
tion of the fluorine is thus expelled, and since the amount of
fluorine in tourmaline is small, the loss on ignition should give
a fair estimate of the amount of water which the mineral con-
tains. Loss on ignition seems to have been very carefully
determined by Rammelsberg, and in those analyses in which

* Berlin Akademie, 1890. Mineralchemie, Zweites Supplement, 1895.
+ Mineral Chemie, Zweites Supplement, p. 284.

110 Penfield and Foote—Composition of Tourmaline.

fluorine is estimated the water is given as equal to the loss on
ignition less eether the fluorine or the silicon fluoride equiva-
lent to it, while when no fluorine was determined a quantity is
assumed, usually one-third to one-half of one per cent, and
this quantity or the silicon fluoride equivalent to it is deducted
from the loss on ignition. In the latter case the assumed
quantity of fluorine never appears in the results as a constitu- |
ent of the mineral. Again, no account seems to have been
taken of the fact that on prolonged ignition the ferrous iron
may become oxidized to ferric, thus causing the water deter-
minations to be too low. In this connection it is interesting to
refer to some results obtained by Riggs, who has given with
many of his analyses both the loss on ignition and the direct
water determinations. When much ferrous oxide is present
the loss on ignition is less than the water, but when this loss is
increased by an amount equivalent to the oxygen which the
iron present can take up in changing from ferrous to ferric
oxide, the results become almost equal. The results, in fact,
would almost do for duplicate determinations, as shown below:

Number: 4... 36 37 388 .41 42 44 45 46 47: 4S) osc Oren
Increase of O. ‘03°06 *Ib” °35- °72_ 1°34°1°58 12 1°32 SO oe
Loss on ign...4°35 4°10 4:09 3°62 3°31 2°30 2°17 2°19 2°41 2°88 2°86 2°69 3°59

4°38 4:16 4:24 3°97 4:03 3°64 3°75 3°71 3°73 3°82 3°17 3°60 4:01
H,0 direct .._4°26 3°90 4:16 3°64 3°49 3°69 3°62 3:49 3°62 3°63 3°60 3°34 3-79

We cannot agree with Rammelsberg that the method adopted
by Riggs for determining water directly gives too high results,
for it has been abundantly proved that when properly executed
it gives exceedingly accurate ones. It is safe to assume, there-
fore, that if the determinations of the loss on ignition as
given by Rammelsberg are corrected for the oxidation of the
iron, the results will furnish fair estimates of the percentages
of water.

Accepting as an established fact that SiO, and B,O, are
always present in tourmaline in the proportion 4:1, Rammels-
berg’s results may be modified by correcting the water deter-
minations and estimating both SiO, and B,O, by difference,
thus distributing the errors of analyses upon two constitu-
ents, rather than upon the B,O, alone. Below are given the
results of applying this method of correction to seven of
Rammelsberg’s analyses made upon varieties of tourmaline
which have subsequently been investigated by either Riggs or
Jannasch and Kalb. ‘There are also given for comparison the
analyses of the other investigators and Rammelsberg’s figures
for $i0,, H,O and loss on ignition. »

a
°
2

xu
2)
o

Penfield and Foote—Composition of Tourmaline. 111
Gouverneur, N.Y., Monroe, Conn., Haddam, Conn., Snarum,
Brownish-
Brown. Black. Black. Black.
Rm.1i. Rgs.53.4 Rm.% Rgs.51. Rm. 11. Regs. 47. Rm. 12. J. &C.56.
Og rs 161 ail 1:10
S §Si0.---[35°76] 37:39 [36-40] 36-41 [34:75] 3495 [34-56] 35-64
iS) 1 B.0,.. [10-434 10°73 [1061] 965 [10:14 992 [10-08] 993
pees ol'3s2) 2-79 31°18 31-27 * - 30°87 ay IGE 30°00 29 41
Fe,03-_- Sees “10 es See 228.2 50 ae 2°90
HeO s... 114 64 407 3°80 8°54 11°87 11°16 6°56
Rin elas Spe i oy Pe pee 09 cee ices
niehy 22-1489 14°09 990" OAT 8°60 4°45 7:94 8°00
@a02... 1°60 2°78 1°81 98 1°33 81 65 1:65
Wa,O0:. 2 1°28 1-72 1:82 2°68 1°60 2°22 1a) tS} 3°03
es pecOi__.° °26 "16 ‘44 21 aid 24 53 "16
= nO. > 3°32 3°83 Bieri arf) 3°44 3°62 3°63 2°94
2) ie ee vohee oA 2s) eee wee Sabet BD oe
100°00 100°42 100°00 99°87 100°00 10035 100°23 101-32
meeIO, . 38°85 39°01 37°50 37°22
S alee. 2°31 2°82 1°81 1°64
Hem o2.°° 3°19 3°32 2°49 2°39
Alabashka, Brazil, Pierrepont, N. Y.,
Black. Green. Black.
Rm.20. J.&C.57. Rm. 32 Rgs.41. J.&C.64. Rm.35. Regs. 50.
Ot 2 i a oe Bren Sea! ae sets "55
3 | BiG; 5 2-0. [so44yy. 354b -_[35°90]17 3691) © 37-05-. [34:98]. 35-61
SeeOqt-= (10-33) 1014 - [10-51] 9°87 9:09 [10-20] 10°15
ee Al,O3)22. 30°41 BPs 37°81 38°13 40°03 27°18 25°29
iO Bert oe 31 em oe ee maa:
Bay 23 15°59 13°42 5°83 3°19 2°36 9°08 8:19
RimOQscs. “b4 a= 1°13 2°22 2°35 = ee Eee
MeO 2s. «1°88 157 "92 04 32 LAS 13°07
Clie rae Ae. gle aes 38 “47 2°91 Bre lt
Was... 102 2°08 2°21 2°70 3°18 1°50 lal
BT) es te “AT 34 42 °28 pirae Dees °20
= yh ae Che mae 1:30 1°61 G0 ee he ae
= 2 6 3°88 3°41 3°57 3°64 3°23 4°02 3°34
24 OU ee "16 "28 “70 14 1°16 erase 27
100°32 =6100°57 100°30 99°42 99783 00:00 -' 99:93
emOa.. . 36:19 38°06 36°64
m+ HO). - Ek] 2°23 3°01
eet ion 2... 2°15 2°92 ?

With the present interpretation of Rammelsberg’s analyses the
ratios of SiO,: B,O, have been assumed as 4:1, while those of
SiO, to the total equivalent of hydrogen atoms in the different
analyses are as follows:

oa 1° No.7
areca, 42198

No. 1F

4:20°5

No. 12
4:20°'06

No. 20
4:19°6

No. 32
4:19°9

No. 35
4:90*2

*Compare the reference to this analysis on p. 115.

112 Penfield and Foote—Composition of Tourmaline.

The agreement between the corrected analyses of Rammels-
berg and those of the other investigators is very remarkable,
and the ratios of SiO, to the total equivalent of hydrogen
atoms approximate closely to 4:20 (the average is 4: 20-1).

Seven complete analyses are also given by Rammelsberg
and to them corrections for water have been applied, while
the errors of the analysis have been distributed upon the
SiO, and B,O,. The results are given below, together with
Rammelsberg’s figures for SiO,, B,O,, H,O and loss on ignition :

No. of Analysis, 2 5 22 24 29 31 33

es SiO, ..-[36°53] [36°37] [34°82] [34:87] [35-49] [86°03] [36-02
SS (| B.03---[10°66] [10°61] [10-16] [10-17] [10°35] [10-51] [10-51}

2 AlsOs =) 32790 33°15 35:46 33:35. 42°63 44050 eee-en

HeO.) 66 2°88 13°17 11:95 eae: : oe 6°38

MnO 259 2 3m" Say 28 1:25 1:94 92 18

MgO_._. 11:79 10°89 1°52 63 39 20 1:88

CaO 125 EWA Pz fess 45 wat ae

Na.O 2°37 1:52 98 1°75 2°60 2°00 2°47

RO) 2) eT Hig 09 “40 68 1:30 47

Pe alii! oe Mls ancet “84 17 1227 72

ces {Ho 3-00 3°81 3:28 4:3] 3°61 3°37 3°65

SOE) is ieee ae 64 aed 4l "82 es 70 55

10027 10000 100717 100°34 100749 10030 100-23

(SiO. _.. 38:09 38°33 36°11 36:22 38°19 38°85 38:46

Spee GbsOs. 2 11S 9°86 11°64 10°65 9:97 9°52 9°13

Bs 1 H,O)2 2.) 92-05 2:34 1:26 D-ait 2°00 2°41 2°31

(len): 992-98 3-49 1°82 2°98 3°61 3°37 2°94

A comparison of the corrected analyses with the figures
given by Rammelsberg indicate that B,O, has been determined
with a good degree of accuracy. His water is always low
and his silica high. The ratios of SiO, to the total equivalent
of hydrogen atoms in the corrected analyses are as follows:

No. 2 No.5 No. 22 No. 24 No. 29 No. 31 No. 33
Ratio 4:20°0 4:20°3 4:20°4 4:20°7 4:21°8 4:21°4 4:20°2

Two of the hydrogen ratios are very high; four of them
approximate closely to 4:20°0; the average is 20°7.

The results obtained by applying similar corrections to

the remaining twenty analyses are given in the table on page
118, together with Rammelsberg’s figures for $i0,, H,O and
loss on ignition. In the last column the ratios of the total
equivalent of hydrogen atoms are given, one quarter of the
corrected $10, being taken as unity.
The ratios of the silica to the hydrogen atoms in these
twenty analyses exhibit a rather wide variation, from 4:18°7
to 4:21°8. The average, however, is 4: 20-4, and the average
for all of his thirty-four analyses is also 4: 20°4, or approxi-
mately 4:20.

113

uné.

of Tourmal

10n O

it

Penfield and Foote—Compos

£-02
8-16
6:06
T-06
8-06
8-81
G.06
€ 1%
L-81
6-81
1-06
6:16
6-16
9-61
G61
6-02
8:16
9-06
1-06
8-02

“O1yey

§6-€ 19-3 91-1
69-€ 67-6 92-8€
86-§ 09:6 PL LE
PE-€ 00-6 ¥6-8E
49-6 06-1 PI-LE
PPE 18-1 O&-8E
T€-6 P9-L 90-9€
1Z-& PES G6-9E
60-6 PLT 6-LE
006 GPI P2-LE
66:6 8-6 L0-L€
99-6 I1-G &¥-9§E
G66 Gh-l 66-9€
0S-6 66-6 08-8
CZ-& GL-6 00-8€
€6:6 €PF-G ILI1-LE
ILE 19-6 O1-LE
0€-€ 08:6 ST-8&
~o-€ P0-€ 19-8
0G-€ 68-6 Gh-LE
‘usr O'H “OIS
e—-+- a
‘[CUISIIG

09-001=61-1
6¢6-00I=01.0
1Z-00L=09-0
I¥-001= 86-0
06-001=L¥-0
FE-00T= 08-0
9€-00I=48.0
L6:001=79-0
61-001 =4¥-0
00-001=" ~~
€T-001L= 1-0
00.001I="" ~~
00:00l == >=
90-001=9T-0
O0O0T == a
Go:00L— =
0:00 =~
00:001I=
G1-00T=9€-0
00:00L= ~~

T¥-0
8P-0
FL-0
G&-0

Ou Oot

L1-@
16.0
VE 0
8&-0
G-0
66-0
89-0
9F 0
G9-0
68-0
0€-0
0€-0
G6-0
86-0
66-0
€P-0
€1-0
LE-0
0€-0

O°

ferro as
€G-1 69-0
OP:G Ses
¥0-G 0€-0
rae co:
Ine erie
9€-T 61-0
Eve 1 OF-0
6-1 0¢.0
€6-1 69-0
F0-6 er ed
98-1 PF-0
T6-L 60-1
61-6 61-0
EFT 1é-1
84-1 08-0
00-6 Gé-1
00-2 10
€L-3 91-0
L3-% 88-0
O'8N O80

19-0
69-1
T¥-0
0€-P
89 1
90-1
81-0
CE-G
69-6
G9-§
6P-€
¥8-€
66-9
LL-9
LG-h,
€P-6
19-6
11-6
9P-01
69-I1

O31

96.0
Se a
19-3 88-1
69-8 &F-9
18-1 69-01
89.2 08.01
ILO OF-LI
09-1 28-21
07-0 © 28-€1
Me FONT
[9-0 99.81
rs SOM
es BBB
89.0 £6.6
te BEA
"Sree aod
at cae COR
60-0 86%
96:0 08-2
a
OUN Oe"

€8-LP
16-87
68-17
61-€&
GI-PE
€G-1€
PE-0€
16-6
66-06
69-TE
98-1€
GL-PE
PV-0€
60-0€
I¥-L€
96-1E
9G-PE
99-05
GG GE
98-0€

Sorry

[48-01] [st-19]
[or-o1] [¥9-ce]
[cc-ot] [60-98]
[stot] [18-¢¢]
[so-or] [or-F]
[1-01] [26-9]
[e6.6] [¢0.7¢]
[o6-6] [96-¢¢]
[oc-ot] [to-9¢]
[sr-o1] [¥6-¢e]
[tr-o1] [¢9-¥¢]
[eo-o1] [se¢-Fe]
[62-6] [x9-¢¢]
[se.ot] [st-¢e]
[z9-01] Lor-9¢]
[st-ot] [ss-r¢]
[90-01] [xF-F¢]
[ep-o1] [9ox-¢e]
[29-01] [6e-9¢]
[ce-o1] [1%-¢¢]
‘OF "OI:
*‘poqoe1109

euezoy
ysuByleys
eq

eq

eqla
ys[ndereg
S1sgsvoipuy
yorquniy
Aoviy, Aoaog
nejlequosue’yT

‘AN ‘qreyed

UBvMIWINI yy
‘HON ‘49109,
eq
plevyyoy 4S
[eporae xy
qeeypoy
‘td ‘SBvxoy,
[BY AO[[1Z
yoojsueqry
“AqBOOT

AM. JOUR. Sor.--Fourta Srrizs, Vou. VII, No. 38,—FEBRUARY, 1899.

8

114 Penfield and Foote—Composition of Tourmaline.

As already stated, Rammelsberg admits that his water deter-
minations are too low; he also states that the formula for a
complex mineral like tourmaline cannot be determined from
a few analyses but must be derived rather from the average of
many determinations. If then it may be assumed that his loss
on ignition, when corrected for the oxidation of the iron, gives
a close approximation to the amount of water, his results
on the whole agree with the formula H,,B,Si,O,,. It may
also be stated that if the analyses are not corrected they lead
to no satisfactory formula.

REVIEW oF hices’ ANALYSES.—Twenty analyses of Ameri-
can tourmalines were made by Riggs, and the ratios derived
from them furnish the very best evidence of the accuracy of
his results. The ratios are as follows:

Total Total
No. SiO. : B20; : hydrogen, No. SiO. : .B,03; : hydrogen.
36. ee OS: tras JOG 46, 4 3 (O°962 52) 2002
37. Avoy SopiOr3. s) 2 20e5 47. 4° °$ ODS 2) 120705
BOs 7845 VS 0:92) wise 925 48, 4: $ sl sO1 See ores
39. Ay a5, OS 04 Hnezay ab Ont 49, Aw tye) T Seca
40. 44) Soe OZ OG couse) LOE 50. 4. $3. 0:98. sae
41. A Fi Se O92) ase ALOE Sl. ~-4— S ROpS ey eo
42. 4S OTT, es La o2. 4 OP OPO A Seer ane
43. A See OO AS ie 2OLO 53. 4: 3 O97, Seen
44, 4 Sa LOSS haie ns 20cZ 54, 4° OOS setase
45. AS Si Os lony ie 4 3) FO seu

20°03 55.

The average of these ratios is 4: 0°95: 19°88, or a very close
approximation to 4:1:24%, which indicates that tourmaline is
derived from the acid H,,B,Si,O,,. It is pointed out by Riggs
that ‘‘the boric acid found invariably falls short of the theory.”
This is generally, though not always, the case, and it is pre-
sumed that this slight defect in the analyses is due to the fact
that it is not always possible to obtain a correct determination
of boric oxide by the Gooch method without a second fusion of
the silicate with sodium carbonate, which Riggs does not men-
tion having made. It is not indicated by the ratios that these
analyses “give as a general tourmaline formula the simple
boro-orthosilicate R,BO,28i0,” suggested by Riggs. The
nearest approach to this is analysis No. 53, in which the ratio
of SiO, to the total hydrogen is 4:18°9. The ratios with few
exceptions show a very close approximation to the rational
numbers 4:1: 20. In eleven cases the numbers for the hydro-
gen ratios vary between the narrow limits 19°8 and 20:2: How
exact the analyses must be in order to yield such ratios may be
best understood when it is known that a difference of one-half
of one per cent in the estimation of either silica or water would

Penfield and Foote—Composition of Tourmaline. 115

change the numbers of the hydrogen ratio + 0°27 in the one
case and + 0°38 in the other. If the silica were one-half of
one per cent high and the water correspondingly low, the effect
upon the ratio would be to change it from 4:1:20 to 4:0-99:
19°35. The evidence is therefore convincing that, with the
exception of analysis No. 53 (brown tourmaline from Gouver-
neur, N. Y.), the analyses of Riggs are very exact, and also
that the material he analyzed was very pure.* Referring to
page 111, it will be observed that analysis No, 53 is the one
which compares most unfavorably with Rammelsberg’s cor-
rected analyses. If the silica in this analysis were 1°25 per
cent lower and the Al,O, correspondingly higher, the agreement
with Rammelsberg’s corrected analysis would be quite satis-
factory, and the ratio would correspond with the theory. Leav-
ing out of consideration this one analysis, which may be con-
sidered as either defective or made upon impure material, the
average of the ratios of Riggs’ analyses becomes 4 : 0°95: 19°91.

The analyses of Riggs were very severely criticized by
Rammelshberg, who characterized tourmaline analysis as “Aeon
Thema fiir Anfdnger,’ but in the light of our present inves-
tigation we find the results very accurate, and it may justly be
said that we are indebted to Riggs for the best series of tourma-
line analyses that has ever been made. In fact, our conclusions
regarding the composition of the mineral might readily have
been deduced from his results alone. :

REVIEW OF THE ANALYSES OF JANNASCH AND KanB.—
Nine analyses were made, from which the following ratios have
been calculated :

Total Total
No. SiO. : 3B.0O; : hydrogen. No. SiOz : B.O3; : hydrogen.
56. co! 27 Rte clams 1 ha 61. Symone ted a 9 ac 5 aio dpe ap
57. Pee ne an es 19°S 62. ae SO. = 7 200
58. ton 0-05, 3. 20°4 63. Pee Org 8 1 LO]
59. Tee Oa. Ss (L838 64. ere as. “s 2OO1
memes = O83: “204° Averape, £* 3 0°92" "3 ~° 19-9

These analyses, like those of Riggs, bear every evidence of
having been made with great precision, and the ratios, with the
single exception of No. 59, approximate closely to 4:1: 20,
thus furnishing additional evidence that the acid from which
all tourmalines are derived is H,,B,Si,O,,.. The analyses do not

*Tn a personal communication from Professor Riggs the following statement
is made concerning the quality of the material investigated by him: “The
material analyzed was of excellent quality, selected with great care. The color-
less, pink and light green varieties were transparent, gem-like crystals, and the
material of the rose-colored, brown and black varieties was, in my opinion,
equally pure.” Dated, Hartford, January 4th, 1899.

116 Penfield and Foote—Composition of Tourmaline.

indicate the general formula R,. BO,.2S810,, proposed by Jan-
nasch and Kalb. Their boric oxide determinations are in all
cases a trifle too low for the theory, but it is believed that the
reason for this is to be sought in imperfections of the method*
for determining this constituent in a complex silicate. The
analyses are excellent, and they rank with those of Riggs,
among the best analyses of tourmaline that have been made.
From our own experience, however, it is very questionable
whether tourmaline contains so much ferric iron as recorded
in some of these analyses (2°90 to 6°68 per cent).

THE ANALYSES OF SCHARIZER.—The ratios of the three

analyses are as follows:
Total

No. SiO. : 3B.O3; : hydrogen.
oD eres teehee 4 3. 50°70 + 2 gees
1) ERR aD MALE en res 4 8 0°76 3 eee
OT omer nee mete 4S. OA OE

By comparing these ratios with the ones which have been pre-
viously considered, it would seem that there are good grounds
for believing that in these analyses the B,O, determinations are
too low, and that the bases have not been determined with
extreme accuracy. The ratios of SiO, to the total hydrogen
atoms in the main substantiate the formula H,,B,Si,0,,.

ANALYSES FROM MISCELLANEOUS SOURCES.—In looking over
the literature a number of analyses have been found which
need to be recorded. They have been made partly for the
purpose of identifying the mineral and partly for the purpose of
determining the character of the tourmaline from special
localities, but none of them have been made for the special
purpose of determining the chemical composition of the spe-
cies. It seems sufficient to give the ratios only, which are as
follows : ;

Total
Locality. Analyst. SiOe: B,0s : hydrogen..

69:+Mt. Bischoff -._--._.-- Sommerland - - -- - 4 3 0°98 : 185
70. Campolongo --------- Engelmann ---- -- 4-3 0°82 3 ee
71. Sysersk, Urals ----- -- Cossa and Arzruni 4 3; 0°89 ; 18°8
72. Montgomery Co., Md. Chatard...-..--- 4 3 0°84 $ 20°4
78. Nevada Co., Cal...-_-- Melville: hse 4.95. Ee 2088
74. Tamaya, Chili .-. _-- Schwarz 2.2/2.2 4.3 102) 9104
75. Straschin, Bohemia... Weisner .--.----- 4 2° 10ug 70m
76. Urulza, Siberia _..... Stchusseff .._-..- 4: 0°85. 3 e0
71.0 %.olax, lndia 3.2 he Chapman Jones.- 4 : 0°87 ; 1971
Average.... 4 3 0°92 3 ig

* Bodewig, Zeitschr. fiir Kryst., viii, p. 211, 1883.

+ 69, 70, 71 and 72 quoted in Dana’s Mineralogy; 73, Bull. U. S. Geolog. Sur-
vey, No. 90, p. 39; 74, Zeitschr. deutsch. Geol. Gesell., xxxix, p. 238; 75, Min. u.
Petr. Mitth., ix, p. 410; 76, Zeitschr. Kryst., xx, p. 93; 77, Min. Mag., xi, p. 61.

Penfield and Foote—Composition of Tourmaline. 117

As indicated by the variations in the ratios, the analysts
apparently have not had sufficient experience to enable them
to deal successfully with such a difficult problem as the tourma-
line analysis. The average of the ratios, however, approximates
to 4:1:20 and thus substantiates our formula.

TITANIUM IN TOURMALINE.—Titanium seems to have been

overlooked in the earlier analyses of tourmaline, but is reported
in several of the analyses of Riggs, Jannasch and Kalb, and
others. The quantity, however, has always been found to be
small, amounting to over one per cent (reckoned as Ti0,) in
only four of the analyses and to over 0°6 per cent in but two
others. It has not been taken into consideration in making
the foregoing calculations, partly because it would exert no
appreciable influence on the final result, but chiefly because it
is uncertain whether the titanium in. this mineral plays the
of a tetravalent element replacing silicon, or of a triva-
ent element replacing aluminium. Although it may not
be possible with the data now at hand to definitely settle
this question, still the analyses furnish some evidence that
the element should be regarded as trivalent. This result
has been reached by considering titanium both as TiO, replac-
ing SiO, and as Ti,O, replacing Al,O,, and comparing the ratios
derived from the four analyses in which the TiO, has been
recorded as over one per cent. The results are as follows:

Total
No. Locality. eae SiO. : B,O; hydrogen.
: ;, 1 GY pemeenie -4 5 0788.5 18°9

oh Lai Penn. Nesieaad titanium _.4 ; 0°91 : 19°6
88 Ti,O, 1°45 per cent...4 : 0°91 : 19°97

TiO, 1:19 per cent.-.4 : 0°95 : 18°5

Ta -- e

538. Peerenven, NX... Neglecting titanium__4 : 0°97 : 18°9
&& vi 0. 1-07. per cent_- 4: 0°97 ¢ 19.2

4 ae TiO, 1°10 per cent..-4 ; 0°93 : 1971
; eat, Se Gal Neglecting titanium__4 : 0:96 : 19°7

‘ Ti,O, 0° 99° per cent...4 ; 0°96 ; 19-9

TiO 1:22 per cent...4 : 0°90 : 18°4

59. Tamatawe ee) oe 99 * 18
et acchland® Kalb Neglecting titanium..4 :; 0°92 ; 18°8
Ti,O, 1°10 percent __..4°; 0°92 ; 19°1

In these four cases it will be observed that the calculations
give the closest approximation to the normal tourmaline ratio
when the titanium is regarded as Ti,O,. Moreover when the
titanium is regarded as TiO, the departure from the normal
ratio is so great that it does not seem probable that this is due
wholly to defects in the analyses. Some very careful and
exact analytical work must be done in order to decide this
question definitely.

118 Penfield and Koote—Composition of Tourmaline.

CONSTITUTION OF TOURMALINE.—The evidence is convine-
ing that all tourmalines are derivatives of a complex boro-
silicic acid H,,B,Si,O,,, and it is believed that further analyses
will not alter this result, although they may furnish important
data concerning the constitution of the acid. All of the hydro-
gen atoms of this acid in tourmaline are not replaced by
metals, for the different varieties have always been found to
contain water, which indicates the presence of hydroxyl. The
ratio of the silica to the hydrogen derived from water
(hydroxyl!) plus the fluorine is not constant, but varies in Riggs’ -
and our analyses from 4:38:14 (pale green tourmaline from
Auburn, Maine) to 4:2°48 (colorless tourmaline from De
Kalb). In all of the analyses in.which water has been esti-
mated directly, a sufficient quantity has been obtained to yield
two hydroxyls in the formula; in only a few cases has the
amount been sufficient to yield three hydroxyls. We are thus
led to believe in the existence of two hydroxyls in all tourma-
lines, and it seems natural to associate them with the two
boron atoms. The acid consequently becomes H,,(B.OH),Si,0.,,..
The small amount of finorine which is found in many
varieties of tourmaline presumably plays the same role as
hydroxyl, or is isomorphous with it, as in the case of topaz, of
chondrodite, and of other minerals containing fluorine and
hydroxyl. The slight excess of hydrogen over and above the
‘two hydroxyls may be regarded as basic hydrogen, which
plays the role of a metal. Such a relation is not unknown in
complex mineral compounds.*

One of the peculiar features of tourmaline is that varying
proportions of metals of different valences and of essentially
different character replace the hydrogens of the acid
H,,(B.OH),Si,O,,. In all cases thus far examined aluminium
predominates and is present in sufficient quantity to replace

more than half the hydrogens. From this it has been inferred |.

that an aluminium-borosilicic acid H, A],(B.OH),Si,O,, 2s char-
acteristic for all varieties of tourmaline. The constitution of
this acid may be expressed graphically as follows:

iO O—H
Ao “(8 On) ote On aeee
0 ee
Anois ie ee

* This Journal, xl, p. 396, 1890.

‘Penfield and Foote—Composition of Tourmaline. 119

It would seem that the mass effect of the complex radical
[A],(B.OH),Si,O,,], which has a valence of nine, is sufficiently
_ pronounced to control or dominate all types of tourmaline.
Thus it apparently makes no difference whether the nine
hydrogens are replaced largely by aluminium and to a slight
extent by alkalies; or largely by magnesium and to a slight
extent by aluminium and alkalies; or to about an equal extent
by aluminium, iron or magnesium and alkalies; the result in
all cases is the mineral tourmaline, with its characteristic crys-
tallization and its peculiar optical and electrical properties.

The following example (compare the ratio derived from the
analysis of the green tourmaline from Haddam Neck, page
107) will illustrate the method ot determining to what extent
the nine hydrogens of the tourmaline acid, H,A1,(B.OH),Si,0,,,
are replaced by metals of different valences: }

Ratio. Total hydrogen
equivalent, or 20H. 2(OH,F). 18H. Dd tae (Ord,

Al,O, 2°327 9327 1-385 “949 - -154—= 6:1 RB!”

FeO -059 )

mn +056! « : ‘ is eh Tal

MgO -oosy ‘169 169 1g 2154 9-t R

CaO -046 J |

Na,O ‘068 asin, " “15 aiets She a | '

Lid oe 178 178 Pier 15412 Be

er sss : Say AVE ANS A

er en 404 308 +096 096 + 154 = 0°6 H.
20)3°078 2)2°770 9)1°385 9-0

154 1385 154

From the ratio of the total hydrogen equivalent, #, (repre-
senting two hydroxyls) are deducted. The remainder, 18 H,
is divided by two, thus determining the ratio of the nine H’s
replaced by Al, in the formula. The excess of the aluminium
or trivalent metal ratio, R’’”’, together with the ratios of the
bivalent metals, R’’, the alkali metals R’ and the excess of
hydrogen, H, represent nine H’s, which are divided among the
different constituents. Thus in the green tourmaline from
Haddam Neck 6:1 hydrogens are replaced by Al (R’”’), 1:1 by
R’”, 1:2 by R’ and there remains 0°6 excess of basic hydrogen.

The analyses of Riggs, Jannasch and Kalb, Scharizer and
Chatard, given on page 555 of Dana’s Mineralogy, together
with our own analyses, practically include all varieties of tour-
maline which have thus far been investigated, and in the fol-
lowing table are given the results of applying the foregoing
method of calculation to them :

120 Penfield and Foote—Composition of Tourmaline.

No Locality. Color.
{| 1|Minas Geraes, Brazil____|Pale-pink -
ge iRuomtords Me. Sy sees Rose 4 5s
an | SiMbragilig Atti hs eee Green) os
S A.| Auburn, Me, 02.2.2 Colorless _-
oa 5 |Minas Geraes, Brazil____,Pale-green
8 6 |Schtttenhofen, Bohemia_|Red _____-
= {| 7/|Haddam Neck, Conn. ---|Pale-green_
a 8 |Barra do Perahy, Brazil..|Green _-__-
x | Oi Rumatordisie ie 2 = aoe Dark-green
<3 | | 10/|Minas Geraes, Brazil... -|Olive-green
4 WT | Actiouray Me cars. Geese Light-green
12 |Schittenhofen, Bohemia _|Blue-green
LT Si Aaa ournse Mie 8 Ss a Dark-green
. . {| 14|Buckworth, Australia. __|Black -~_-__
oS 15! Parise yhiic wee.) lcite at Blaék 22
=, | 16 |Auburn, Me ..-.-.--.-- Blacks. J. ¢
5 8 17 |Alabashka, Russia.----- Black 2c
& © 4/18 |/Mursinka, Russia_.:--.--|Black__..-
os 19 |Schtittenhofen, Bohemia _|Blue-black _
AS 20) Picdrapblanca <2 es 22 Black en:
oo 21 |Montgomery Co, Md.-_-|Green (Cr)
| | 22 |Minas Geraes, Brazil_.__|Black ____-
. (| 23" Stony omt NCL ae. Black es
Bh ae | -24\Olebpran 2 Sao )o ee ae Bideley. 2 ==
oe | 25. |Tamatawies 2 oF oo ee Blaekec a.
35 26 |Haddam, Conn. ._-_-.. YUBA Ck as 2s
o _ 27 |Snarum, Norway... --- Blaek
ee 111| 28 Orford Ne Eimee boned Dark-br’wn
= H | 29 | Monroe: Conny’). a. ieus Dark-br’wn
| | 30 |Nantic Gulf, Baffin’s Ld._|Black _-. --
ay -({ | 31: |DeKalb, Nove -- 87. Colorless - -
2s 4 32 | De KealiowNepien. ¥ ought eee. Colorless _-
ot 5 33 |Gouverneur. N Y _ ___. |Brown ___-
So [ 34 | Hamre, Nery oe) mee Brown ----
“gs i 30_|Pierrepont, IN. Yo. 22 ae== Black 222%

Analyst.

Riggs, 36
Jann. & Kalb, 64
Riggs, 38
Bice 30.42 ee
Scharizer, 67 ._-
Authors
Jann. & Kalb, 63
Riggs, 42
Riggs, 41
Riggs, 40
Scharizer, 66___
Riggs, 43
Jann. & Kalb, 62
Riggs,
Riggs, 45
Jann & Kalb, 57

Jann. & Kalb, 60

Scharizer. 65__-
Jann. & Kalb, 58
Chatacd, 72___-
Riggs, 40222

Jann. & Kaib. 56
Riggs, 52
Riggs, 51
Riggs, 49
Authors

Riggs, 54
Riggs, 53
Riggs, 55

Riggs, 50__ 222

2

QA ONATSARR ARR wWwWwWKEWWWWNHEHOHHHHOSOOHOS
WOW NAATAAH- HP BRORADNABRNHNSONNHESOCANGHUEHHE MADEN Y

et ee SF a Maple ape Sarees are elon OM alt
WE BRWAARDAMANAAAMANATAACAMRTHDY HWE RYH wWwOH SwD

i
fe

tt

SSeS SS 2 SSS SOLS SSS See eS SoS SS So SS SSS eo
T HDOWMNTOTANW OH BP TTTAOORODADAYP EDP AONAwWATH THEO]

It will be observed that the extent to which the nine hydro-
gens of the acid H,A1,(B.OH),Si,O,, are replaced by metals, is

very variable.

The trivalent metal, R’”’, is chiefly aluminium,

and the extent to which the hydrogens are replaced by it

ranges from 6:7 to

16.

The bivalent metals represented by

R” are chiefly iron and magnesium, and the extent to which
the hydrogens are replaced by them ranges from 0-2 to 6:3.

In general R” and R” are reciprocal.

There is always some

hydrogen replaced by the alkali metals R’, and some basic

hydrogen.

The results are arranged in the table according to

a decrease in the replacement of the hydrogens by R’”’, and
according to this arrangement the different varieties of tour-
maline fall into natural groups.

The first 18 examples are characterized by containing an
appreciable quantity of lithium, while in the others none, or

Penfield and Foote—Composition of Tourmaline. 121

not more than traces of this element have been found; these
may therefore be designated as Lithia Tourmalines, which
form a natural group. In this group R’ is higher than in
other varieties, while H is somewhat higher than the general
average; R’” is very high and R” correspondingly low. This
variety of tourmaline has its particular mode of occurrence,
being found in pegmatite veins associated with quartz, albite,
microcline, orthoclase, muscovite and lepidolite. The material
is generally delicately colored and often transparent and of
gem-like quality. It is difficultly fusible when heated before
the blowpipe, fusibility=5—5$.

At the opposite extreme, 31-34, are varieties which may be
designated as Magnesia Tourmalines. In these RK’, which is
chiefly magnesium, is very high and R’”’ correspondingly low,
while R’ reaches its lowest limit, 0°2 to 0-4. These varieties
are light-colored and at times of gem-like quality. They are
easily fusible before the blowpipe, fusibility=3. With these
No. 35 should be associated, for it differs only in containing
considerable iron which is isomorphous with the magnesium,
hence the color of this tourmaline is black. The last five are
alike in their mode of occurrence. They have probably been
formed in limestones containing magnesium by the contact
action of intruded igneous masses during the pneumatolitic
period, when such masses were giving off heated aqueous
vapors containing boracic acid and fluorine compounds. They
are found in coarse crystalline limestone, associated with
graphite, phlogopite, pyroxene, amphibole, scapolite and apa-
tite. Contact metamorphisms of this nature have recently
been described by Lacroix.*

The intermediate varieties (Nos. 14 to 30) with the excep-
tion of No. 21 are black or dark brown, owing to the presence
of iron. These are the ordinary tourmalines found in granites,
gneisses and schists, and sometimes in pegmatite veins inti-
mately associated with lithia tourmaline, as at Auburn and
Paris, Maine. They too have probably resulted from the min-
eralizing action of heated aqueous vapors containing boracic
acid and fluorine compounds, given off during the pneumato-
litic period of cooling igneous rocks. A contact metamorph-
ism of this character, attended by the formation of tourmaline
vein stone, has been very carefully described by Hawes.t+

Nos. 14 to 22 (No. 21 excepted) are characterized by con-
taining iron and only a little magnesia, hence they may be
designated as ron Tourmalines ; Nos. 23 to 30 contain both

* Le Granites des Pyrénées et ses Phenoménes de Contact, Bull. Carte Géolo-
gique de France, No. 64, Tome x, p. 54, 1898.

+The Albany Granite, New Hampshire, and its Contact Phenomena, this
Journal, xxi, p 21, 1881.

122 Penfield and Foote—Composition of Tourmaline.

magnesia and iron, and are hence designated as Magnesia-Iron
Tourmalines. These two groups, however, grade into one
another. The fusibility of these intermediate groups of tour-
malines varies from 4°5 to 8,. and decreases as the amount of
iron and magnesia increase. |

Although in this series of thirty-five analyses there are pro-
nounced groups or types of tourmaline which may be recog-
nized, nowhere in the series do the ratios of R’”, R’”’, R’ and
H approximate so closely to rational numbers that a definite
formula for any one type can be instituted. Riggs and Jan-
nish and Kalb, however, have given special formulas for
different types of tourmaline (pages 100 and 101) based upon
multiples of the acid H,,(B.OH),Si,0,,.. Deducting from
their formulas appropriate multiples of the aluminium-boro-
silicic radical [ Al,(Bb.OH),S$i,O,,], it is found that the nine vari-
able hydrogens of the tourmaline acid are replaced by metals
of different valences in the following proportions:

RZ Re R/ H

tare ies. uggs. oo eee 70) 0°03 a7

I, Lithia tourmaline --- Jann. and Kalb. 6:0) 1st
: Riges . 2) se ae 5:0. 2°56. a0 areamirg

I. Iron tourmaline. --.- Jann: and Kalb. ..5°0) s-UmeGeaannns

III. Iron-magnesia tour. Jann. and Kalb._-4:0 4:0 0°77 0°

Concerning the formulas for lithia tourmaline, Riggs’ cor-
responds closely to No. 1, and Jannasch and Kalb’s to Nos. 7,
8, 9 and 10 of our series (compare the table on p. 120), but
neither of these complicated formulas furnishes a satisfactory
expression for this type as a whole. Similar statements might
be made concerning their formulas for iron tourmalines and
for iron-magnesia tourmalines. The endeavor to express the
composition of tourmaline or of one of its types by a definite
formula may be compared to the attempt to express the com-
position of the dark varieties of sphalerite by a formula. Thus
Zn,,Fe,S,, would correspond to sphalerite, containing about 50°5
per cent of zinc and 15:7 per cent of iron, but zine and iron
are isomorphous and can mutually replace one another in
sphalerite, and a variety containing less iron would have to be
expressed by a different formula. When, however, we under-
stand the isomorphous relations existing between zine and
iron, and express the composition by RS, where R=Zn and
Fe, the composition becomes very much simplified.

In tourmaline we have an isomorphous relation of a very
peculiar nature, for in the acid H,AI,(B.OH),Si,O,, the nine
hydrogens may be replaced to a large extent either by the
trivalent metal aluminium or by the bivalent metals magnesium

Penfield and Foote—Composition of Tourmaline. 123

and iron without any decided change in crystalline form. This
leads to the consideration of a certain phase of isomorphism
which, as it seems to us, has not been considered with suffi-
cient care, namely, the mass effect of complex radicals in
influencing or controlling crystallization. Thus in their sim-
ple salts we do not regard sodium and potassium as isomorph-
ous with calcium, but in some complex silicates, zeolites for
example, we recognize Na, and K, as isomorphous with, or at
least capable of replacing, calcium, barium and strontium. In
some of the phosphates, dickinsonite and fillowite for example,
we have sodium (Na,) playing the same role as calcium, man-
ganese and iron in replacing the hydrogens of phosphorie acid.
In the garnet-sodalite group we have minerals with isometric
erystallization, to which the following formulas have been
assigned :*

3

Fe, Al(SiO,), Almandite

3

(Cl— Al)Na,Al,(SiO,), Sodalite
(NaSO,—Al)Na,Al,(SiO,), Noselite
(NaSO,— Al)(Na,, Ca), Al,(SiO,), Haiiynite
(NaS, — Al) Na,Al,(SiO,), Lazurite
[(OH, F, Cl),.Al],Al,(SiO,), Zunyite

When it is taken into consideration that isometric erystalliza-
tion is exceptional in the group of silicates, we are led to
believe that the sexivalent radical [A1],(SiO,),], by virtue of its
mass effect, controls or dominates the crystallization of these
minerals. Not only are they isometric, but, with the excep-
tion of zunyite, which is tetrahedral, they all crystallize com-
monly in dodecahedrons. In sodalite, noselite and lazurite,
such unlike constituents as chlorine, and the univalent sulphate
and polysulphide radicals (NaSO,)’ and (NaS,)’ play the same
part in the complex molecules. It is moreover probable that
these unlike constituents are isomorphous in the sense that
they can mutually replace one another, for Brégger and Bick-
strom have described homogeneous material containing the
lazurite, haiiynite and sodalite molecules, while appreciable
quantities of chlorine are almost always found in noselite and
haiiynite, thus indicating the presence of the isomorphous
sodalite molecule. It is to be expected that the larger and
more complex the radical the more potent will be its mass
effect in controlling or determining the crystal form. Thus in
sodalite, noselite and haiiynite the- radical is very large,
ore a) (510,), |, RNa, or Ca,.

Tourmaline, it would seem, furnishes an example somewhat
analogous to that presented by the garnet-sodalite group. In

5) } f j
Ca,Al,(SiO,), Grossularite | Garnet

* Brogger and Backstrom, Zeitschr. fiir Kryst., xviii, p. 209, 1890.

124 Penfield and Foote—Composition of Tourmaline.

this acid H,A1l,(B.OH),Si,O,, the mass effect of the complex
radical [A1,(B.OH),Si,O,,] is so great, or, the role played by
the replacement of the nine hydrogens is so subordinate, that
the bivalent elements, Fe, Mg, Mn and Ca, and to a slight
extent the alkalies, Li, Na and K, can replace aluminium as
isomorphous constituents. This conclusion in some respects is
analogous to that reached by Rammelsberg, who stated in 1870
that all tourmalines were derived from an acid H,SiO,, in
which the six hydrogens were replaced in varying propor-
tions by R’,, R”,, Al, and B, Applying Rammelsberg’s idea
to the acid H,,(B.OH),Si,O,,, all varieties of tourmalines may
be regarded as mixtures of the molecules

R’,,(B.OH),Si,0,,, R’ =Li, Na, K and H
R"»}(B.OH),Si,0,,, BR’ =Fe, Mg, Mn and Ca
R’” (B.OH),Si,O R’’=Al, Cr and small amounts of Fe and Ti.

It seems more logical and satisfactory, however, to consider
all varieties of tourmalines as salts of the acid H,A1,(B.OH),
$i,0,,, in which the complex aluminium-borosilicie acid radical
[ Al,(B.OH),Si,0,,] exerts a mass effect by virtue of which the
remaining hydrogens may be replaced by metals of essentially
different character without bringing about any pronounced
change in crystalline form. |

199

Note concerning the detection of ferrous and ferric iron in Sili-
cates; by S. L. Penfield and H. W. Foote.

In the preliminary work connected with the study of the
methods of the tourmaline analysis, considerable time was
devoted to ferrous and ferric iron determinations. The ordi-
nary method of dissolving in a mixture of hydrofluoric and sul-
phuric acids and titrating with potassium permanganate can
not be applied readily to tourmaline because it is extremely
difficult to dissolve the mineral. Riggs overcame this difficulty
by using very fine powder and heating under pressure in a
heavy platinum crucible provided with a lid which was held
down by clamps. Jannasch and Kalb succeeded in decompos-
ing the exceedingly fine, elutriated powder by heating in sealed
glass tubes with a mixture of sulphuric and hydrofluoric acids,
but the method was far from satisfactory. We have employed
fusion with borax. In order to test the method, silicates con-
taining varying proportions of ferrous and ferric iron were
first carefully standardized by the method described by Pratt.*
The powdered silicate and borax glass were then fused in a
platinum boat in a combustion tube, through which a current
of either nitrogen or carbon dioxide gas was conducted. These

* This Journal, xlvili, p. 149, 1894.

Penfield and Foote—Composition of Tourmaline. 125

gases were first freed from all traces of oxygen by conducting
them over a roll of heated bright copper gauze at one end of
the tube. Decomposition is thus readily effected without any
appreciable oxidation. After fusion the material can be dis-
solved in dilute sulphuric acid and titrated with potassium per-
manganate. With small quantities of iron the method gives
very satisfactory results, but ferric iron suffers a slight reduc-
tion during fusion, hence the method can not be recommended
as a quantitative one when accuracy is required. The amount
of reduction is nearly proportional to the amount of ferric iron,
and when about 6 per cent of Fe,O, is present the error
amounted to about 0°3 percent. Itis certain, however, that very
satisfactory determinations of ferrous and ferric iron might be
obtained by means of this method, by applying a correction
which could be determined by experimenting with some arti-
ficial mixture of silicates containing approximately the same
amount of ferric oxide as the mineral. It is difficult to under-
stand the cause of this reduction, which was first noted by
Rammelsberg,* and afterwards carefully studied by Suida.t

Fusion with borax furnishes, however, a most excellent
means for decomposing insoluble silicates, when qualitative
tests for ferrous and ferric iron are desired. The application
of the method is as follows: The mineral and some borax glass
are heated in a closed-tube over a Bunsen-burner flame until
decomposition is complete, or nearly so. Water is then
dropped upon the hot glass in order to crack it, and, when cold,
the end of the tube is broken up and transferred to a test-tube.
About 3° of dilute hydrochloric acid are then added, and the
acid is boiled vigorously until the material is dissolved, when
about 5° of cold water are added. The liquid, after being
divided into two portions, may then be tested with potassium
ferricyanide for ferrous iron, and with either ammonium thio-
cyanate or potassium ferrocyanide for ferric iron. The results
are very satisfactory and decisive, and although there is a
tendency for ferric iron to undergo slight reduction during
fusion, and also a tendency for ferrous iron to undergo slight
oxidation during the manipulation, these defects are so trifling
that practically they may be disregarded.

Mineralogical-Petrographical Laboratory,
Sheffield Scientific Schoo!, January, 1899.

* Zeitschr. der deutsch. geolog. Gesell., xxiv, p. 69, 1872.
+ Min. Mittheilung, Tschermak, 1876, p. 175.

126 ZH. A. Pilsbry—Lrttoral Mollusks from Patagonia.

ArT. XIII.—Zittoral Mollusks from Cape Fairweather,
Patagonia; by Henry A. Pitssry. (With Plate L)

In the course of his paleontological explorations in Pata-
gonia in the interest of Princeton University, Mr. J. B.
Hatcher found time to make a small gathering of marine shells,
at the locality named above. The scarcity of information upon
the fauna of the east coast of South America makes it desir-
able to place the facts gained from even small collections on
record, to the end that the limits of the faunal provinces of
this coast may be more exactly defined, and their characteristics
more fully exposed.

The fauna at the point in question is typically Magellanie,
with but little admixture of types from the Argentine fauna
of La Plata region. Vacella, Photinula and T rophon, Plazi-
phora and littoral Brachiopoda, all occurring in a gathering so
small as this, at once indicate the western and not northern
origin of this mollusk population. Lullia alone is more
northern in distribution, this being its most southern outpost.

The addition of the genus Sphenia to the Magellanic fauna
is interesting and unexpected.

The figures on Plate I are three-fourths natural size.

Mouricip 2.
Trophon Gerversianus Pallus.
Trophon Gerversianus Phillipianus Dkr.
Trophon fasciculatus Hombron and Jacquinot.
Trophon textiliosus Hombr. and Jacq. Pl. I, fig. 4.

Shell fusiform, white or slightly buff tinted, with moderately
dilated body-whorl and short oblique or subvertical anterior
channel. Whorls about 6, convex, separated by a deep suture,
the last dilated toward the lip. Sculpture of numerous sub-
regular rounded spiral cords (28-34 on body-whorl, 8-11 on
‘penultimate whorl), separated by much narrower linear
grooves, and numerous vertical folds, which are rather closely
placed, separated by intervals narrower than themselves, are
somewhat arcuate in harmony with the lines of growth and
become obsolete upon the base. There are about 20 (18-21)
folds upon the penultimate whorl, pretty regular and equal; but
upon the latter part of the last whorl they become irreoular or
in part obsolete. Aperture ample, over half the total length
of the shell, the outer lip somewhat expanding and faintly
crenulated ; ‘columellar lip with asmooth white callous, reflexed
over and closing the axial chink; siphonal fasciole well
developed.

Alt. 36, diam. 23 mm.; alt. of aperture 22 mm.
44 39 66 21 ce ce 66 91 66

H. A. Pilsbry—Littoral Mollusks from Patagonia. 127

The original description and figures were from a young
shell. The species has not hitherto been recognized, probably
owing to this circumstance. Mature specimens collected by
Mr. Hatcher are here described and illustrated. The figures
referred to decolor Phil. in the Voyage au Pol Sud. (Zool., pl.
25, f. 6-8), look like adults of textzliosus.

T. textrliosus has been reckoned a synonym of 7. Gerver-
sianus by Tryon (Man. Conch. IL and IIL), but this is only
another example of the “lumping” practised by him in deal-
ing with Zrophon.

_It differs conspicuously from most other Cape Horn Tro-
phons in its less rude sculpture and symmetrical contours.
T. decolor Phil. has a much shorter anterior canal, purple
interior and fewer spirals. In Z. cancellinus Ph. the lip is
thickened and dentate within. TZ. crispus Gld. differs in
sculpture; and Z. liratus Couth., which is apparently the
most nearly allied species, has the anterior canal longer, and
there are fewer spiral grooves. Specimens are in the collection
of the Academy of Natural Sciences of Philadelphia from
“Santa Cruz River, Patagonia, Wm. Bell.”

BUCCINID &.
Euthria plumbea Phil.

Enthria sp. undet. Like plumbea, but finely spirally
striated throughout. All of the specimens are beach-worn and
not suitable for description.

Bullia globulosa Kiener.

V OLUTID &.
Scaphella Kerussaci Don.
Scaphella ancilla Sol.
CALYP ZETRID &.

Trochita corrugata Reeve. Varies toward forms in which
the radial ribs almost disappear.

TROCHIDA.

Photinula ceerulescens King.
Photinula teeniata Wood.

Photinula pruinosa Mab. and Roch. Quite distinct from
the longer-known species in its Gibbula-like contour and wide,
flattened, lusterless columella.

PAaTELLIDA.

Nacella enea magellanica Gmel.
Nacella enea deaurata Gmel.

Mopatiip”.
Plaxiphora setigera King.

128 H. A. Pilsbry—Littoral Mollusks from Patagonia.

PECTENID ZA.
Pecten patagonicus King.

Myip&.
Sphenia Hatcheri n. sp. PI. I, figs. 5, 6.

Shell cretaceous and dull, stout, rather strong, moderately
inflated, with broad, somewhat truncated, closed anterior,
and narrow, abruptly truncate and widely gaping posterior
end. Beaks small, in close contact and eroded at the tips,
nearly median, being one mm. nearer the posterior than the
anterior end; projecting somewhat above. Dorsal margin
highest anterior to the beaks, thence slopmg downward and
posteriorly; anterior end widely subtruncate, arcuate; ventral
margin slightly arcuate. Sculpture, fine, nearly regular con-
centric grooves separating wider intervals on the anterior por-
tion of shell, disappearing posteriorly, where there are only
some rude, irregular wrinkles at unequal intervals. Face of
the chondrophore in right valve directed ventrally ; that of the
left valve partially erect. Interior white or stained, porcellan-
ous, the adductor scars distinct, pallial line and very broad
posterior sinus indistinct. Length 17, altitude 9-5, diam. 7-2™™.
A single specimen collected. The genus has not before been
reported from this region to my knowledge.

MactTRID&.

Mactra (Mactra s. str.) Jousseaumi Mabille and Rochebrune.
PINT, figs. 15 °2, 3.

The specimen figured, one of several collected, agrees with
Mabille and Rochebrune’s short and unsatisfactory description*
as far as that goes, except in size; their specimen measuring,
length 45, alt. 34, diam. 18™™. The proportions however are
similar, the figured specimen measuring 67, 48, 28". The sub-
triangular form, perfectly straight slope of the valve margins
in front of the beaks, and obtuse siphonal extremity, as well as
the details of hinge structure, are sufficiently shown in the
figure. In external texture it is like an old Spisula solidissima,
but the thin, light yellowish-olive cuticle adheres in small
ragged patches.

Mytinip&.

Mytilus magellanicus Gmel.

Mytilus sp. A single small specimen.

Mytilus edulis I.

TEREBRATELLID&.

Terebratella dorsata Gmel.

Academy of Natural Sciences, Philadelphia.

* Miss. Sci. du Cap Horn, Mollusques, p. 106.

Starkweather— Thermodynamic Relations for Steam. 129

Arr. XIV.—The Thermodynamic Relations for Steam; by
G. P. STARKWEATHER.

AmMonG the various thermodynamic relations for a gas that
between pressure, volume and temperature, the ‘equation of
condition’ is one of the most important. To apply to the
case of steam such equations, either empirical, or based on
assumptions hardly justifiable, have been formed by Zeuner,*
Schmidt,t Ritter,t and Antoine.§ Of these Zeuner’s and
Schmidt’s satisfied the saturation line well, but they, as well as
Antoine’s, had the fatal defect that the constant which deter-

mines the limiting value of a for very great volumes was

assumed incorrectly.
Van der Waals’ celebrated pamphlet in 1873 gave a rational
form to the equation of condition, and Clausius’ generalization |
of the same,
; aE F(T)
ia Gee

has been for the most part adopted for various substances.
Equations based upon this generally have the aim of satisfying
not only the gaseous state but also the liquid. A relation
between a and §, and also the value of f(T) at the critical
point, follow from the fact that at that point not only does this
equation hold, but also its first and second partial derivatives
with respect to v. KR is fixed by the chemical composition of
the substance, while /(T) isso determined as to satisfy the
corresponding saturation pressures and temperatures by
Clausius’ method.§{ The value of a (hence @, from the rela-
tion mentioned above) is so chosen as to satisfy the water line
at some point.

The best application of this method to steam has been given
by Van Laar,** who has taken for A(T) the form adopted by
Sarrautt for carbonic acid, ey", « and y being constants. On
applying Clausius’ method of determining the corresponding
saturation pressures and temperatures to the equation he has
found a good agreement.

* Zeitschr. des Ver. Deutsch. Ingen., xi, p. 42, 1867.

: er. des Ver. Deutsch. Ingen., xi, pp. 649, 771, 1867, and Prag Abhandl.,
' + Wied. Ann., iii, p. 447, 1878.

; Comptes Rendus, cx, p. 632, 1890.

Wied. Ann, xiv, pp. 279, 692, 1881.

§] Wied. Ann., xiv, p. 692, 1881.

** Zeitschr. fir Phys. Chem., xi, p. 433, 1893.
++ Comptes Rendus,:ci, pp. 941, 994, 1145, 1885.

Am. Jour. So1.—Fourra Series, Vou. VII, No. 38.—FrEBRvARY, 1899.
9

130 Starkweather—Thermodynamic Relations for Steam.

But Van Laar’s equation does not satisfy the saturation vol-
umes. The writer has shown in a previous paper* that the
saturation volumes as calculated from Regnault’s latent heats
are entirely out of accord with those directly measured by
Battellist and has given reasons for preferring the former.
From these Van Laar’s equation deviates 2°5 per cent. Even
should we assume Battelli’s experiments to be correct instead
of Regnault’s the deviations would still exceed 1 per cent.

Were it certain that the equation were necessarily of
Clausius’ form, holding for the liquid as well as the vapor, the
method would be exact, and we should simply have to say so
much the worse for the experiments. But all we know con-
cerning the last term of the second member is that at ge

volumes at any given temperature it must be sensibly — ate

has been shown by Van der Waals. If we replace eh UF
pe’ (+8)

; Ler ee : 1
by some other function of » which is sensibly wi at large volumes,

the application of Clausius’ method becomes very complicated,
if not, indeed, impossible. The failure of Van Laar’s vn)

f(T)

(v+py
not sufficiently correct for the water line, at least.
Abandoning the attempt to satisfy the water line throughout,

I(T)

we might still consider 5

to satisfy the facts would seem to show that the form

; as sufficiently accurate for it in

the vincinity of the critical point, the relations for that point
still holding true. The equation would thus apply to steam
for all states. In this case, if we give the equation Clausias’
general form, his method for determining f(T) does not apply,
and we are free to choose it so as to satisfy the saturation
volumes. This has apparently never been done. The writer
has attempted it, but without success. 7(T) would necessarily
decrease very rapidly with eats temperature. Thus at
100° C., 200°C. and 365°C. (the critical temparature) f(T)
-would have the respective values 59° 5, 21°5 and 13°5. More-
over, the great fall from 100° to 200° is out of accord with the
change from 200° to 865°. This seems to indicate that the
type of formula selected will not satisfy the facts.

There remains the following problem, to obtain equations
which shall satisfy the experiments within ordinary limits.
Such equations would suffice for most purposes, and moreover
the writer hopes to draw from them some theoretical conclu-

* This Journal, Jan. 1899, p. 1
+ Mem. R. Ace. Se. di Torino ai 43, 1893.

Starkweather— Thermodynamic felations for Steam. 131

sions of interest. A p-v-Z equation should be of the form

p=" _s(rj6).

(Yt!

In this R must have its theoretical value deduced from the
chemical composition of water, while a, on account of the
extreme incompressibility of water, should have a value not
much less than 0-001, v being in cubic meters per kilogram.

Moreover, ¢(v) must be sensibly - for large values of v.

The writer has formed the following equation, to hold for
volumes not much less than 0-12 cubic meters per kilogram.
At such volumes it should not be used when the pressure much
exceeds 11700™, although for larger volumes it may be applied
to higher pressures.

RT A
v—a_ Tv? (vt +y)

—

If p is expressed in millimeters and wv in cubic meters per
kilogram the constants are
log R = 0°539990 a = 0°0008
log A = 4632127 y = 1°:20484

The value of R is that given by Van Laar,* modified by the
fact that he places the absolute zero at —273°-2 instead of
—273°°7, as assumed here.

How well the equation agrees with the saturation line is
shown by the following table. In the first column is given
the temperature absolute, in the second the pressure, p,, calcu-
lated from the formula by means of the saturation volumes,
in the third the pressure according to Regnault, p,, and in the
fourth the fractional deviation, d, of p, from pz.

Ly Pe pr d it pe p d
813-7 55°11 54:91 +:°0036|403'7 2026°4 2030°3 —-0019
893-7 92°37 91°98 +:0042/413°7 2712°2 2717-6 —-0020
333°7 149°34 148°79 +:0037|423°7 3573°3 3581:°2 —:0022
343°7 233°66 233°08 +:0025|433°7 4641-7 4651°6 —-0021
853°7 355°04 354°62 +:0012/443°7 5950°9 5961°7 —-0018
363°7 525°64 525°39 +:0005/453°7 75381 7546-4 —-0011

373°7 760°01 760°00 0 463°7 9442°7 9442°7 0
383°7 1074°6 1075°4 —‘0008/473°7 11700 11689 + ‘0009
393°7 1489°5 1491°3 —'0012

With the exception of those near 323°-7 the variations are
insignificant, but the change in sign at 463°°7 shows that the

* Zeitschr. fir Phys. Chem., xi, p. 433, 1893.

132 . Starkweather—Thermodynamic Relations for Steam.

equation will not hold for much smaller volumes. The larger
deviations near 323°-7 are not due to the form chosen for the
equation, for at such large volumes the gas deviates only
shghtly from the law of Boyle and Charles. A. sufficient
explanation for these deviations les in the values of p and

ae Regnault’s formula for which is incorrect at low pressures,
due to the fact that he did not recognize the sudden change in
dp 4 go
a at O°.

pand T being given, the formula is not convenient for
obtaining v, but if an approximate value of v be found from
Boyle and Charles’ law, p obtained, and then w corrected by
the assumption

This matter will be considered later.

dlogp = —d log»,
a second correction will generally be sufficient.
ee

If the factor 7 3 correct, the term v(v?+r) must be sen-
sibly so, for it satisfies the volumes on the saturation line and
approaches the value v’ for large values of v. To justify the

1 :
factor 7p recourse would naturally be had to volumetric
experiments on superheated steam. Since it has been shown
in the previous paper, however, that such experiments do not
appear reliable, a more roundabout method is necessitated to
justify the factor.

We now proceed to obtain the equations for entropy, 7, and
energy, «. The equations for energy formed by previous
writers have all been based on the assumption that either the
specific heat at constant pressure or that at constant volume is
constant, or on another assumption equally unjustifiable, which
will be mentioned later. There will be needed the quantity

ee la
and since
de = Tdyn — pdv,
we have
dy = —ndT — pdv.
From this equation follow |

oy

Sv_|r

ES ee
ya Pa

Substituting the value of » from our equation of condition in
the first of these, and integrating, there results

=-—p and

Starkweather— Thermodynamic Relations for Steam. 183

2A 2A pity
= lees (oa) ee oe
v ) Be (v—a) Tyot * Ty ea Ute)

where F(T) is some unknown function of T. It should be
observed that differences of yr at constant temperature formed
by means of this equation are more accurate than the value of
p found by the equation of condition, for an integral cannot
have a greater percentage of error than the integrand.

Making use of the equation

Binge
Yj Sar 5ST ;
there is obtained
2 4
n= Rlog, (v—a) — = bots Igoe eed --F'(T).

oie fe)
5 ie vr Pies We ms yt
ae ;
It is at once evident that here the factor pin the equation for p

plays an important role, and this equation, together with the
one to be obtained presently for the energy, may have a great

il Reis
error at small volumes, unless the factor — is justified.

T
From the equation
f= W+Tn

we find

4A 4A ve +y

e= — — — lo + f(T

yTv? yT 3 ot I( )

where

FUL) =F (T)—FP Ae).

F(T), except for an additive constant, is the kinetic energy, or
rather such part of it as depends on the temperature alone.

A number of writers have performed this same integration
with their p-v-T equation, and have immediately assumed
that 7(T) is linear in the temperature. We are now in position
to show that this assumption is contrary to the experimental
data at hand by determining /(T) within certain limits. For,
from the equation for e, we have
b vi+ y

a
T) =e+—-+ 1
F(T) Set + aH h08,.

where, if eis expressed in kilogrammeters and v in cubic meters
per kilogram, the constants are given by

log a = 6:286713 y = 1:20484,
log 6 = 6°567998

1384 Starkweather—Thermodynamic Relations for Steam.

Now « and v are known on the saturation line for values of T
from 273°-7 to 473°°7. For there eek

e= JH— p(v—w)

where H is the total heat, J the mechanical equivalent, and w
the volume of water at the pressure and temperature under
consideration. Consequently /(T) is found for this range.

The values given for the energy on the saturation line by
the last equation are untrustworthy at low temperatures, since
there v is rather uncertain owing to the error in Regnault’s
formula for corresponding saturation pressures and tempera-
tures. But from the equation for 7, since for the region below
100° C. w is constantly 0°001, there is obtained

RT(y=001) (AGE Oe
» — :0008 Tv2(vt+y) _

p(v—w) =

In the first term of the second member, since w is large, any
error in v cancels out. The last term is so small for the region
where serious errors in » may occur that the volumes calcula-
ted from the latent heats may be used. ‘This also is true con-
cerning the values of the potential energy; any slight errors
in the volumes at low temperatures will not affect things.

In this manner e¢ has been calculated for the saturation line,
then 7(T), and the following table obtained. In the first
column is given the absolute temperature, in the second the
corresponding value of f(T), and in the third, under /(T), the
average rate of increase of /(T) for each ten degrees.

T f(T) f(T) T FL) Ty) T FL) ae)

273°7 242870 343°7 252135 413°7 261862

141°5 126°4 146°6
283°7 244285 353°7. 253399 423°7 262828

137°3 127°0 151°2
293°7 245658 363°7, 254669 433°7 264340

133°4 12771 155°7
303°7 246992 3873°7 255940 443°7 265897

131°4 128°5 159°5
313°7 248306 383°7 257225 453°7 267492

128°7 133°4 163°3
323°7 249593 393°7 258559 463°7 269125

128:°0 137°3 165°2
333°7 250873 403°7 259932 473°7 270777

126°2 143°0

Leaving out the lowest temperatures, which will be con-
sidered presently, we have the very remarkable fact that /(T),
instead of being constant, as is generally supposed, is an

Starkweather—Thermodynamic Relations for Steam. 135

increasing function of the temperature. Thiesen* in a recent
paper has shown that the total energy, expressed as a function
of vw and T, is not linear in T, but it has never before been
shown, so far as the present writer can discover, that the
kinetic energy is not linear, although Massieut has given a
complicated proof that the specific heat at constant pressure
increases with the temperature; this, indeed, was contempo-
raneously done by Weyrauch.t MRegnault§ has found experi-
mentally the same fact for carbonic acid gas.

This fact that 7(T) is an increasing function of T is so
important that it is well to speak of the proof more in detail.
We have from page 133

SJ (T)=e—7,

e, and 7, denoting respectively the values of the total and
potential energies on the saturation line. At 100° C., 150° C.
and 200° C. the values of ¢, are 254757, 259571 and 264336
respectively, and if the last number were increased by 49, e,
would be linear in T. On the other hand the corresponding
values of —7r, are 1183, 3257 and 6441, the zero of potential
energy being that of indefinitely rare gas, and the last exceeds
a linear formula for 7, by 1110. Now it is impossible that by
any change of p-v-T equation, hence a change of the form
of a, the number 1110 can be reduced to 49.

We have next a remarkable corroboration of our formule, in
: ee ;
jp in the last term of the p-v-T
equation, by means of Regnault’s experimental determination
of the specific heat at constant pressure. At temperature
403°-7, pressure 760™", the volume is found by the p-v-T
formula to be 1°8018 cubic meters per kilogram. We have
then, if Q, is the quantity of heat necessary to turn one kilo-
gram of water at 0° C. into this state at constant pressure,

«= JQ, —p(v—w) = JQ, — 18609.

particular of the factor

But from the formula for e on page 183, since we know (page

134) £(403°7)= 259932, we have e=258909, hence JQ,=277518.
Similarly at temperature 473°°7, pressure 760™", we find the

volume to be 2°1314. We thus obtain in the same way

JQ, = 292035,

Q, being the supply of heat necessary to turn one kilogram of
water at 0° C. into steam of this second state, at constant
pressure.

* Wied. Ann., 1897, No. 13, p. 329.

+ Mem. des Savants Etrangers, vol. xxii, p. 58, 1876.

t Zeitschr. des Ver. Deutsch. Ing., xx, pp. 1, 71, 1876.

§ Mem. of the Institute of France, vol. xxvi, pp. 128, 129.

136 Starkweather—Thermodynamic felations for Steam.

Accordingly, we find for the mean specific heat at atmo-

spheric pressure from 403°°7 to 4738°°7

Q,—Q, Perc
OF ae a0

or only 0-9 per cent greater than Regnault’s mean specific
heat at atmospheric pressure from about 401°°7 to 490°-7,
namely 0°4805. When one considers that this deviation could
be cancelled by a diminution of less than 0-05-per cent in the
latent heat at 200° C., which is far within the limits of error,
the essential corroboration of the formule is evident.

Gray * has attempted to show that Regnault’s experiments
on the specific heat of superheated steam give a value of -
0°3787. He does this by comparing the total heat at 100° C.
with the quantities of heat obtained in the experiments on
superheated steam. But this is not allowable, as Regnault ¢
himself distinctly states, and he would not even compare the
heats in different series of experiments. In fact if the latter
be compared some absurd results are obtained.

We have now to express f(T) by a formula. To get as high
a determination as possible the writer has taken Regnault’s
second determination of the mean specific heat at constant
pressure, namely that from 138°C. to 226°C. at atmospheric
pressure it is 0-48111. Assuming this to be true from 130° to
230°, we obtain the quantity of heat necessary to heat the
steam at atmospheric pressure between those temperatures.
Adding this to JQ, found on page 135 we obtain the quantity
of heat necessary to heat water at atmospheric pressure from
0° C. into steam at 230° C., JQ,. At the latter state the volume
is 2°2718, so subtracting p (2°2708) from JQ, we find the energy
at this state. From the energy formula /(T) is found to be
275278.

It will be noticed that at low temperatures 7’(T) apparently
changes sign. This is entirely dependent on the formula
chosen for total heats at those temperatures. In the preceding
article, in forming the expression for the total heats from 0° C.

2

to 100° C., it was assumed that since was negative above

at

100°, it was probably so below. Under this assumption we see
f"(T) to change sign. On the other hand, if we make f”(T) of

one sign throughout, namely positive must change sign.

d
? dt?
Now it is far more probable that the latter does so, for it
concerns only a particular set of states, namely those on the
* Phil. Mag: xiii. Sat, Loon:
+ Mem. of the Institute of France, vol. xxvi, pp. 165, 166.

Starkweather— Thermodynamic Relations for Steam. 187

saturation line, whereas /’(T) refers to any states whatever.
The writer has therefore assumed 7/(T) to be represented in
the range considered, so far as the accuracy of the experiments
permits us, by the formula

J(T) =a4+6T+cT

Determining a, 6 and ¢ from the values of 7(T) at 273°°7,
373°°7 and 5038°°7, there are obtained.

a= 215126 lgb=1:902449 log c= 2894845

How well /(T) is represented by this formula is shown by
the following set of values calculated from it, which should be
compared with those given on page 134. In order to properly
represent the inaccuracies of the formula, in the third column
are given the fractional deviations in the latent heats which it
would be necessary to assume in order to make the formula
absolutely correct.

i f(T) d ea FCP) d eget F(T) d
273°7 242870 0 353°7 253200 —-0009|433°7 264536 +:0009
283°7 244107 — +0008 363°7 254562 —°0004443°7 266023 +:0006
293°7 245358 —°0012/873°7 255940 0 '453°7 267527 +:0002
303°7 246626 —°0015 383°7 257334 +:0005463°7 269045 —-0004
313°7 247910 —-°0017/393°7 258743 +:0008473°7 270580 —-:0010
323°7 249209 —-:0016'403°7 260167 +:°0011/503°7 275278 0
333°7 250524 —-0014/413°7 261608 +-0011

343°7 251854 —*0013/423°7 263064 +:0011

It will be noted that the negative deviations near 313°7 bring
the latent heats nearer Griffiths’ determinations. It is probable
that the latent heats thus calculated from /(T) as expressed by
the formula are from 0° C. to 100°C. nearer the truth than
those obtained from the formula for H given in the preceding
article.

Theoretically it would be very gratifying if /(T), which is
the specific heat at constant volume for very large volumes,
should at low temperatures approach either 3R or $R. These
are respectively 141°44 and 117°37. That the first is not
approached is evident from the numbers on page 134.
Whether or not the second is, is not conclusively shown; the
total heats are not accurately enough known at low temper-
atures to determine this. Since 3 R would be 165-01, it looks
as though in the range say from —50° C. to +200° C. a change
in the molecular deportment of the steam occurs, such that the
number of degrees of freedom of the molecule increases, and
the specific heat at constant volume changes from §R to 7 R.
Such molecular change might explain why a difficulty exists
in finding a p-v-T formula to satisfy both water and steam.

1388 Starkweather—Thermodynamic felations for Steam.

From the relation
J(T) = F(T)—TF(T)
there follows
(2) = — TFT)

from which F(T) and F(T) are determined. The formule for
n and wv on page 133 become accordingly

b’ ghee aa
=a' log,,(v— _ +d’ log,,T+eT+K’
0 iy dl esata g,Tte T+
ih ; BONG vit y
y=-—a Tlog,, (va)— Tot T log,, yt We

dT log,,T — e” T—— T+"

where if 7 is in kilogrammeters per degree, Ww in kilogram-
meters, and v in cubic meters per kilogram,

log a’ = 2°035657 log 6’ = 5°985683
log c’ = 6'266968 log d’ = 2'264662
log e' = 1195875 log e” = 2:050449
RS S11 922 K? > -S=eQ5 1G:

These formule, with that for the energy, have the same lim-
itations as to pressure and volume as the p-v-T formula, and
in addition can be used only when T is not far above 503°°7 or
far below 273°°7.

The value of K’ was determined so as to make the entropy
correct for the saturation line at 373°°7; the remaining quanti-
ties follow from others previously determined. How well the
formula for 7 represents the saturation line is shown by
the following table, in the third column of which are given
the fractional changes in the latent heats necessary to make
the formula for 7 correct. These should be the same as those
given on page 187, but there are differences, very slight,
however.

T U7) d iT n d mt n d
273°7 935°4 +:°0011)| 343°7 788°2 —-°0011/413°7 702°5 +°0010
283°7 908°5 0 353°7 773°1 —°0010|423°7 693°3 +0011
293°7 883°9 —:‘0007 | 363°7 759°2 — 0006 |433°7 684-4 +°0008
303°7 861°4 —:0011/373°7 746°3 0 443°7 676°3 +:°0008
313°7 840°7 —:0015|883'7 734:°2 +°0005|453°7 668°5 +:0002
323°7 821°7 —:0015|393°7 722°9 +°0005|463°7 661°3 —-°0002
333°7 804°3 —:0013/403°7 712°4 +:°0009)473°7 654°5 —°0005

It has been remarked that Regnault’s formula for correspond-
ing pressures and temperatures is incorrect at low temperatures.
Also Wiebe* has taken exception to his determinations near

* Zeitschr, fir Instrumentenk., xiii, p. 329, 1893.

Starkweather—Thermodynamic frelations for Steam. 139

80°C. There will now be developed a formula for p from the
latent heats.
We have from Clausius for the saturation line

from which follows
dp; dlogp. dh
paw aT sc 2% Tp (ps0):

Let us denote this by D. We have already obtained p (v—w)
with great accuracy (page 134) for temperatures below 100° C.
For L we take the values from the assumed equation for /(T),
which seem to be the best obtainable. Thus D is found for
various values of T.

Next the writer has assumed that D is represented by

A B C
Dima tp ope:

This is Rankine’s form, with the addition of the term =
Determining A, B and C from the values of D at 273°°7, 3238°°7
and 3738°°7, there are found

log A = 3°539061 log B= 5:826374 log C = 7°556878

How well this represents D is shown by the following com-
parison. Under D is given the value as first found from the
latent heats, and under D’ that according to the formula.

T D jE} Al, D D’
273°7 0°0724625 0°0724626 | 333°7 0°0462057 0°0462064
283°7 0°0667980 0:°0667857|343°7 0°0432180 0°0432188
293-7 0°0617441 0°0617298)353°7 0°0405067 0°0405049
303°7 0°0572187 0°0572100/363°7 0:0380399 0°0380322
313°7 0°0531516 0°0531546|373°7 0:0357738 0:0357738
323°7 0°0495038 0:0495037

Integrating, there is obtained

AS. » BR ee?
log, P= — ar aw tat K
where
log A’ = 3'176848 log B’ = 5'163131
log C’ = 6°717544 K = 7'844259

K is determined by making p equal to 760™" when T is 373°-7.
In the following table the values of p according to this formula
are given in the second column, in the third are Regnault’s

140 Starkweather—Thermodynamic Relations for Steam.

pressures, ~,, in the fourth the fractional deviations of the
former from the latter, and in the fifth a few determinations
at low temperatures recently made by Juhlin.* |

T p Pane d Ds
Tae T A696 (A 006 4-618
983°7 9°277 9-165 ‘012 " 9°949
293°7 17°632 17°391 014 17°46
303°7 31°9438 31°548 013
Bay, 55°449 54:906 008
395-7 1 392-610 91°980 007
333°7 149-40 148°79 006
343°47 233°59 233-08 "005
353°7 354:°94 354°62 003
363°7 525°56 5 25°39 0008
Sia 760 760 Bo

The variations are considerable, but for those at low tempera-
tures.an explanation has already been given, and in the middle
of the scale they are not more than could be ascribed to Reg-
nault’s not correcting his thermometer to the air-thermometer.
Above 100° C. Regnault’s gives two formulas, one for tempera-
tures according to the air-thermometer, the other for tempera-
tures by his mercurial thermometer; below 100° this is not
done, the inference being that corrections were not made.

Now Regnault remarkst+ that at about 50° C. the air-thermom-
eter differs from the mercurial by two-tenths of a degree,
which would be sufficient to account for the deviations in that
region. The pressures are on the same side of Regnault’s as
those of Juhlin, and agree well from 80° to 100° with those of
Wiebe.

In the following table are given values of various quantities for
the saturation line from 0° C. to 100° C. on the basis of the formulee
developed in this paper, and from 100° to 200° from the for-
mulee adopted in the preceding article. They represent, in the
writer’s opinion, the best determinations possible with the data
at hand. In the first column is given the temperature Centi-
grade, in the second the heat of the liquid, in the third the
total heat, in the fourth the latent heat, in the fifth the pressure,
in the sixth the volume, in the seventh the entropy of water,
N.», and in the eighth and ninth the entropy and energy of dry
saturated steam, 7, and e,.

* Beiblaetter zu Wied. Ann., xviii, p. 736.
+ Comptes Rendus, Ixix, p. 884, 1869.

Starkweather—Thermodynamic felations for Steam, 141.

t h i § L p v Nw Ns Eg)

0 0 —598°9 5989 4°626 204:97 0 935°4 242851
10 10:0 602°8 592°8 9:277 105°91 15°55 9085 244072
20 20:0 606°8 586 8 17-632 57°653 303 883°9 245297
30 30:0 610°7 580°7 31°943 32°861 44°55 8614 246524
40 40:0 614°6 5746 55°449 19-518 58°4 840°7 247749
50 50:0 618°5 068°5 92°610 -12°032 iad 821-7 248966
60 60:0 622°3 562°3. +149°40 7°6666 84:°7 8043 250168
70 70°0 626°9 556-0 233°59 5°0333 97°3 788°2 251353
80 80°1 629°7 549°6 354°94 3°3953 109°7 717371 252515
90 90:2 633°3 543°) 525°56 2a4re OMe (5952 253652
100 100°4 636°7 536°3 760 1:6587 13375 746°3. 254757
110 110°5 639°8 529°3 1075°4 ise 144'9 Tas 9) 6255724
120 120°6 642°9 522°3 1491°3 0°88099 15671 722°6 256691
130 130°7 645°9 515°2 2030°3 065948 1670... 711-9 257649
140 140°9 6489 508°0 2717°6 0°50152 177°6 702°0 258615
150 151°1 651°9 500°8 3581°2 038707 188°0 692°%7 259571
160 1613 6548 4935 46516 030278 1981 684:0 260517
170 1716 657°7 486°1 5961°%7 0°23982 20871 675°9 261485
180 181°9 660°5 4786 7546°4 0°19216 2179 668°4 262440
_ 190 192°2 663 3 AQNl:71l 9442-7 0°15566 2275 661°4 263394
200 202°5 666°0 463°5 11689: 0:12739' 2369 -654°7 264336

The unit here for quantities of heat is the specific heat at
15°, pressures are in millimeters of mercury, volumes in cubic
meters per kilogram, energy in kilogrammeters (at Paris) per
kilogram, and entropy in kilogrammeters per kilogram per
degree.

There is a gratifying confirmation of the formule given by
Dieterici’s experimental determination* of the volume at 0° C.,
which was 204°68 cubic meters. This differs from the value
given in the table by only a little over one-tenth of one per
cent. When one considers the roundabout way in which the
latter was found, the agreement is quite remarkable. Dieterici
considers his determination to be correct within one-half of one
per cent.

There are one or two interesting points which can be brought
forward concerning the deportment of steam over ice. At
0° C. that portion of the energy due to the terms in the
formula containing v is only nineteen kilogrammeters, which
is entirely negligible compared with the latent heat. Hence
for steam over ice we can set

ex f(T}.
Now if G isthe latent heat of ice, and K its specific heat, we have
«= JL—JK (273:7—T)—JG—p (v—w).
Hence
JL—JK(273-7—T)—JG—p (v—w)—f(T) = 0.
* Wied, Ann., xxxviii, p. 1, 1889.

142 Starkweather— Thermodynamic Relations for Steam.

Below 273°°7 we can set
pv = p(v—w) = RT.
Then, since
_ Te(v—)}

d dp: , dlo
JL=Te-) R= = RT ae

we have

RT? pe —JK (273-7 -T) —JG@—RT—/(T) = 4

Now first suppose that determinations of L below 0° C. are
made, or that those above 0° C. are so accurate that an expres-
sion for /(T) is obtained for that region which will bear extra-

d |

polation below 0° C. Then a as
between p and T determined. Conversely, if the relation
between p and T below 0° C. is known, 7 (T) is at once deter-
mined, and thence the latent heats.

is known and a relation

A. EF. Verrill— New Actinians. 143

Art. XV.— Descriptions of imperfectly known and new
Actinians, with critical notes on other species, III; by
A. E. Verritt. Brief Contributions to Zoology from the
Museum of Yale College, No. LX.

Family Paractipa Hertwig, op. cit., 1882. Andres, op. cit., p.
255, 1884. McMaurrich, Proc. U. 8. Nat. Mus., xvi, p. 160,
J1., 1893.

Paractide + Sideractide Daniel., 1890.

Paractide + Actinostolide Carlgren, Kongl. Svenska. Akad. Handl., xxv, II, pp.
64, 137, 1893.

Entacmeous actinaria provided with a mesodermal sphincter

muscle, but destitute of acontia. Mesenteries usually numer-
ous, nearly always with 12 or more pairs perfect, the num-
ber and arrangement not always truly hexamerous. The
perfect and larger imperfect mesenteries are usually fertile.
Column-wall is usnally smooth, sometimes with submarginal
plications and solid ridges; rarely with low wart-like elevations
or small verrucee, sometimes capable of attaching sand, ete.
No actinobranchs. No acrorhagi. No disk-tubercles.
_ The mesoglea of the wall is usually tough and elastic, thin
or thick, often parchment-like or subfibrous, in deep-sea forms
often thick and coriaceous. Base with an adhesive disk, which is
sometimes amplexicaul, embracing permanently the stems of
gorgonie, etc. Mouth with two siphonoglyphs and numerous
lateral folds. ‘Tentacles simple, usually numerous and retrac-
tile. Margin generally capable of involution.

This family appears likely to become an extensive one when
all the species, whose places in the system are still doubtful,
shall have been studied anatomically. At present there is no
practical and sure way to distinguish the Sagartiade@ from this
family except by the presence of acontia in the latter, and
this is not always a satisfactory test* with preserved specimens.
However, the thin, smooth, tough, wall is often a fairly good
indication of this family, as is also the large number of perfect

* All the acontia may be ejected and lost by violent contractions, when certain
Sagartie are placed in preservatives, or they may be easily decomposed in poorly
preserved specimens. So they are often not to be found in specimens known to
have had them in life. On this account it is doubtful whether several of the
genera described by Danielssen, 1890, as destitute of acontia, belong to Paractide
or Sagartiade. Of these Stelidiactis, Altantactis, Anthosactis, Kyathactis, and
Korenia are described as having pores or cinclides in the wall (the last has them
also in the disk). These genera may, therefore, be sagartians that have lost
their acontia. The first three have only 6 pairs of perfect and ‘sterile mesen-
teries, as is often the case in Sagartiade ; the fourth has 20 perfect pairs; the
last has 24 perfect pairs ; the base is amplexicaul. His Sideractis is octamerous,
with 16 perfect pairs of very thin mesenteries, partly fertile; tentacles not

retractile, 8 inner ones large; sphincter mesodermal. It probably belongs with
Paractide,

144 A. EF. Verrill—New Actinians.

and fertile mesenteries, which are usually thin and have feeble
and diffuse longitudinal muscles.

This family now appears to include Paractis, Stomphia,
Actinostola, Actinernus, Actinolobopsis,* Pycnanthus, ? Sider-
actis, Cymbactis (near Stomphia), Tealidiwm, Paranthus,
? Par actinia, Kadosactis, Phelliopsis,t nov. (type P. Pana-
mensis V. ’69), Alloactis, nov. (type A., excavata Hert.),
Laphactis, nov., Synanthus Ver., Ammophilactis, nov. (type
A., rapiformis Les.).

The last named genus is remarkable for its small base and
for having minute adhesive suckers near the margin. It agrees
with Paranthus in living buried in sand and in its elongated
form.

Raphactis, gen. nov. Type &. nitida V.

Paractides having a short column, with thin wall and sub-
marginal longitudinal plications, with raised folds or solid ridges
_ of mesogloea. Sphincter muscle enlarged distally in a thick-
ened portion of the wall, above which the thin wall can be
involuted ; below this fold the wall may besmooth, or wrinkled,
or slightly warty, but without suckers.

Base either broadly expanded, with thin edges, or amplexi-
canl and embracing permanently stems of gorgonie, ete., and
in this case the opposite edges of the same or of different
individuals unite by sutures.

Tentacles numerous, rather stout, in several rows, retractile.
Mesenteries numerous, 12 or more pairs perfect and fertile;
some imperfect ones may also be fertile. Stomodeeum short.
This genus lacks the pores of the disk and wall described in
Korenia. Synanthus Ver. is closely allied in form and habit ;
it has a smooth wall and only six pairs of perfect, but fertile,
mesenteries.

Raphactis nitida, sp. nov. Figures 18, 22.

Base large, either broadly expanded, with thin edges, or else
clasping stems of sponges, hydroids, etc., and in the latter case
the opposite lobes may unite together in a close suture; it
secretes a tough thin cuticle. Column in contraction low and
broad, most of its surface smooth and white in alcohol, often
somewhat glossy or, in others, with a parchment-like appear-
ance. The infolded summit is strongly plicated. Toward the
contracted summit there is a much thickened fold, usually cov-
ered with a series of unequal longitudinally convergent, low,
rounded, solid ridges, which vary in number and size, and are

*This new name is proposed for Actinoloba Hert., 1882 (non Blainy,, 1834.)
The type is A. reticulata (Dana, sp.) of Patagonia, etc.) .

+ This name is proposed for those species of paractids that have been neeetibed

as Phelliew, on account of the adherent cuticle. They have usually 12 or more:
pairs of perfect, fertile, and very muscular mesenteries.

A. EF. Verrill—New Actinians. 145

often lacking or concealed in the smaller examples; in the
larger ones about 24 longer, alternating with shorter. Ten-
tacles of moderate size, thick, tapered, blunt, crowded in five
or six rows; in the larger specimens sometimes more than 144
in number; the inner are much longer and stonter than the
outer ones. The walls of the tentacles are thick, and uniform.
In section the column-wall is generally thin, but tough;
toward the collar it becomes gradually much thickened and
very firm. Sphincter muscle mesoglceal, imbedded in the
thick collar, large, diffuse, club-shaped distally. Mouth large,
with two broad shallow siphonoglyphs and about 12 to 14 dis-
tinct folds on each side. The siphonoglyphs may lie in the
direction of the short axis of the body, or the reverse. Mes-
enteries regularly hexamerous, in more than four cycles, the
fifth usually more or less incomplete ; 12 pairs are perfect and
fertile and in the larger specimens 16 pairs may join the
stomodeum near its upper end; many of the smaller imper-
fect ones of the fourth cycle are also fertile; those of the
fifth cycle are very small except close to the disk and basal
membranes where they become larger and some bear small
gonads close to the disk. No acontia were found. The
stomodzeum is very short and broad, strongly plicated within.
The gonads are small; those of the perfect mesenteries are
close to the base. Septal foramina of large size occur at about
mid-height of the column. Diameter of column of the larger
examples, contracted in alcohol, -85 inch (20-22™"); of base
1:5 inch (386™"); height.of column ‘75 inch (18™”).

Color in life was not noted.

Station 1048, U.S. Fish Com., Str. Albatross, in 130 fathoms
off E. coast U. States, latitude of Delaware Bay.

This species has the habit common to many genera and
species of deep-sea Paractide and Sagartiade of clasping
with its base the stalks of hydroids, gorgoniz, sponges, etc.
When the opposite lobes come in contact they unite in a
suture.* So, likewise, do the edges of the bases of adjacent
individuals, as shown in fig. 18, A, B, C, and fig. 22.

* Hertwig made this habit a generic and family character, in the case of
Amphianthus and Stephanactis, giving it, undoubtedly, an exaggerated importance,
for the same species may live either in this way or attached by a broad base to
a shell or stone, as I have repeatedly observed in Actinauge Verrillii McMur.
(= nodosa Ver.) and other species. It is even doubtful whether it should be of
generic value in any case. Slephanactis was used by me in 1868 for a genus
near Discosoma, consequently I propose for the group thus named by Hertwig the
new name. Sfephanauge. It is near Actinauge and Hormathia. The type,
Stephanauge abyssicola (Moseley, 1877) is identical with my Actinauge nexilis
(1883) and came from the same region. S tuberceulata (Hert.) is very similar, if
not the same, but it came from the Pacific Ocean, as did S. hyalonematis (McMur.,
1883), which last is very imperfectly known. The structure of this genus will be

discussed later It differs from Actinauge chiefly in having the thick wall smcoth
below the fold, and perforated by distinct cinclides.

Am. Jour. Sci.—Fourtu Series, Vou. VII. No. 38.—FEBRUARY, 1899.
10

146 A. EF. Verril—New Actinians.

WN GAA
GF A ANS ye Ran Ve .
ALA TS a
ange een
=e mee
Res
Bae ay
areas!
Sy BE
ia ig
i
sa ef
SSae7
i?

Explanation of cuts.

Hig. 16. Paractis nivea (Less.). Peru, x 14.

Fig. 17. Raphactis Caribea Ver. on a gorgonian, nat. size, front and side view,
contracted. West Indies, 124 fathoms. :

Fig. 18. Faphactis nitida Ver., three examples, A, B, ©, natural size, united
together by sutures; contracted in alcohol.

Fig. 19. Sagartia (Psammactis) modesta Ver., + nat. size, from life, New Haven.

Fig. 20. Viatrix globulifera D. and M., 1860. x3. Bermuda.

All the drawings are by A. H. Verrill.

Erratum—In No. II of this series, p. 42, I erroneously referred Actinia cruen-
tata Dana to Bunodactis, overlooking the anatomical description of MeMurrich
(Proc. U. 8. Nat. Mus., xvi, p. 150, 1893), under the name Condylactis cruentaia.
It should, I now believe, be called Actinoides cruentata, as it agrees closely with
the typical species of the latter genus. Page 45, Bwnodella stelloides. Since, the
publication of No. II, I have received from Mr. J. E. Duerden, of Jamaica, larger
examples of this species. On dissection they were found to have 24 pairs of per-
fect and fertile mesenteries, thus agreeing with Bunodactis. The genus Bunodella
must, therefore, be dropped, and this species be called Bunodactis stelloides. More-
over the name Bunodella had already been used by Matthews for a genus of
Hurypteroidea.

L. C. Jones—Estimation af Boric Acid. 147

Art. XVI.—A Volumetric Method for the Estimation of
Boric Acid; by Louis CLEVELAND JONES.

{Contributions from the Kent Chemical Laboratory of Yale University -—LXXVIIL ]

WHEN boric acid and mannite are mixed in solution, a
peculiar compound of strongly acid properties is the result.
This compound decomposes carbonates, and its acid taste
is comparable to that of citric acid, much stronger than that
of boric acid alone. Magnanini* has found that the pro-
duct of such a mixture of boric acid and mannite solutions
shows greater electrical conductivity and a lower freezing
point than a similar molecular solution of either substance
alone. Other polyatomic alcohols (but all to a less degree
than mannite) and some organic acids show this peculiar prop-
erty of combining chemically with boric acid to imerease its
acid qualities.t Of this reaction between boric acid and the
polyatomic alcohols, Thomson,t Barthe,$ and Jérgensen| have
taken advantage to develop methods for the volumetric esti-
mation of boric acid. Glycerine is used to form a combination
with boric acid, sufficiently acidic to give an acid reaction
when used with a sensitive indicator and make possible its
titration with an alkaline solution. Honig and Spitz4] show
that in the method of Jérgensen a very large amount of
glycerine must be used to prevent the appearance of the indi-
cation of alkalinity with phenolphthalein before all the boric
acid is neutralized according to the following equation:
29NaOH + B,0,=2NaO0B0+H,O—; that in the presence of
carbonates the solution must be boiled to decompose bicarbo-
nates and the escape of boric acid by volatilization prevented by
the use of a return condenser, and that silica must be removed
by the process of Berzelius, and the solution then neutralized
by the use of methylorange before a titration of the boric acid
can be made.

Vadam,** for the estimation of boric acid in butter, makes
use of mannite, which, as he finds, gives sharper indication
with litmus than glycerine. According to this process, the
solution to be analyzed for boric acid is neutralized by the use
of litmus and a solution of sodium hydroxide. Mannite
(1-2 grm.) is then added, bringing about an acid reaction with
the boric acid present in free condition. The solution is then
titrated to alkalinity by sodium hydroxide.

* Gaz. Chim., xx, 428-440, xxi, 134-145.

+ Klein, J. Pharm. Chim., 4, vol. xxviii; Lambert, Comp. Rend., eviii, 1016-1017.

J.5.C.1., xv, 432. § J. Pharm, Chim,, xxix, 163.

f Zettecne, f. Angew. Chem., 1897, 5.

“| Zeitschr. f. Angew. Chem. (1896), 549.
** J, Pharm. Chim. (6), viii, 109-111.

148 L. C. Jones—Estimation of Boric Acid.

None of the above methods with glycerine have, in my
experience, given anything but comparatively crude results.
The weak acidic properties of boric acid, the interference (and
difficulty of removal) of carbondioxide with all organie indi-
cators sufficiently delicate to be used with boric acid, and
indeed, the procuring of a standard alkali containing no car-
bonate, together with the supposed detrimental influence of
silica and the lack of a convenient method for its removal,
have made the process of Gooch,* which involves distillation
and weighing with calcinm oxide, the only means (though
requiring long time and exceeding care) in use for the accu-
rate separation and estimation of boric acid. Recently sodium
tungstate has been recommended from this laboratory? as a
substitute for calcium oxide to retain the distilled boric acid.
The entire process, however, is one of the most exacting in
analytical chemistry, and for this reason a convenient, rapid
and at the same time accurate method for the estimation of
boron is especially desirable. The first step toward the
development of such a process must be the convenient prepa-
ration and the accurate estimation of the standard solution of
alkali to be used for neutralizing the boric acid. This has
been found to be easily accomplished by the process recom-
mended by Kiisler.{ This observer, in an extensive investiga-
tion of the analytical methods for the volumetric estimation of
alkalic and alkali carbonates in solution, finds that both phenol-
phthalein and methylorange are appreciably sensitive to
carbonic acid, but when this interfering agent is removed by
precipitation with barium chloride according to the process of
Winkler,§ the remaining free alkali may be estimated with
great accuracy by phenolphthalein and decinormal hydrochloric
acid.

Obviously, if the difficulties dependent upon the action of
carbon dioxide can be obviated, and if the acidity of the boric
acid can be increased to such an extent that a sufficiently sen-
sitive indicator will give with accuracy the neutralization
point with free alkali, and if the alkali and stronger acid can
be combined while boric acid alone remains free, then it
should be possible to estimate boric acid volumetrically. Ex-
periment has shown that barium chloride removes carbon
dioxide, and that mannite makes a combination with boric
acid strongly acidic to phenolphthalein.

* Amer. Chem. Jour., ix, 23-33; Moissan, Comp. Rend., cxvi, 1087: Kraut,
Zeitschr. f, Anal. Chem., xxxvi, 165; Montemartini, Gaz. Chim. Ital., xxviii, 1,
344,

+ Gooch and Jones, this Journal, vii, 34.

t Zeitschr. f. Anorg. Chem, xiii, 124-150.
S Massanalyse.

L. C. Jones—Estimation of Boric Acid. 149

To obtain the boric acid alone in free condition many
attempts have been made. Gladding,* Thaddeefft and Rosen-
bladtt have isolated the boric acid by distillation with methyl-
aleohol and a non-volatile acid. Many indicators theoretically
insensible to free boric acid have been used to indicate the
neutralization of the stronger acids. H6énig and Spitz,§ and
Thomson,| use methylorange, Morse and burton,{ tropaeolin
00, while Vadam** makes use of litmus. All these indica-
tors, however, have been found by experiment to be more or
less affected by boric acid in solution. On the other hand, I
have found in the well-known reaction according to which a .
stronger acid liberates regularly iodine from a mixture of
iodide and iodate, the solution of this difficulty. If both the
iodide and iodate are in excess of the acid the entire amount
of free acid will be neutralized and the corresponding amount
of iodine liberated according to the following equation :

5KI+K103 +6HCl=6KCl+3H,0 +3I,.

This liberated iodine may be removed by sodium thiosulphate
and a solution obtained which is absolutely neutral containing
only neutral salts, potassium iodide, iodate and tetrathionate.
The statements made by P. Georgevictt and Furry,tt that boric
acid present in moderate amount in solution has not the
shghtest action on a mixture of iodide and iodate, have been
experimentally verified. Therefore when this acid is liberated
by an excess of a stronger acid, and the iodine set free de-
stroyed by thiosulphate, it remains free in solution to be titra-
ted in any convenient manner possible. :
Following along the lines suggested by the above reac-
tions, a volumetric process for the estimation of boric acid
has been developed. For a basis of the investigations, a
standard solution of boric acid was prepared by dissolving
in a liter of water about eight grm. of carefully weighed
anhydrous boric oxide. This anhydrous boric oxide was pre-
pared from the several times recrystallized hydrous boric
acid by long-continued fusion over a blast lamp. A solution
of approximately = sodium hydroxide was prepared from the
ordinary sodium hydroxide of the laboratory. The free alkali

* Jour. Am. Chem. Soc., iv, 568.

+ Zeitschr. f Anal. Chem., xxxvi (9), 568.
Zeitschr. f. Anal. Chem, xxvi, 18.

§ Zeitschr. f. Anorg. Chem. (18). 549.

| J.S. 0. LL. xv, 432.

4] Am. Chem. Jour., x, 154.

** J, Pharm. Chim. (6), viii, 109-111.

++ J. Prac. Chem., xxxviii, 118.

tt Am. Chem. Jour, vi, 341.

150 L. CO. Jones—Lstimation of Boric Acid.

in this solution was estimated by the process of Winkler recom-
mended by Kiisler. The acid used to make this estimation
was hydrochloric, standardised by silver nitrate.

The full method for the estimation of boric acid as finally
elaborated is as follows: The solution is made clearly acid to
litmus by hydrochlori¢ acid and 5°™ of asolution (10%) of barium
chloride added. An amount of iodate and iodide of potassium
sufficient to liberate an amount of iodine at least equivalent to
the excess of hydrochloric acid in the acidified solution is
mixed with starch in a separate beaker, and the iodine, which is
usually thrown out by this mixture, is just bleached by a dilute
solution of thiosulphate.

To the now neutral solution of iodide and iodate a single
drop of the solution to be analyzed is transferred by a glass
rod. Ifa blue coloration is developed, the solution is acidic
with hydrochloric acid, and all the borie acid is in free condi-
tion. The amount of iodide and iodate used depends upon
the acidity of the solution containing boric acid. Usually
10°™ of a 25 per cent solution of iodide and the same amount
of a saturated solution of iodate is sufficient. Any larger ex-
cess of hydrochloric acid should be neutralized by sodium
hydroxide before the iodide and iodate mixture is added.
After the addition of the iodide and iodate solution, containing
starch, to the boric acid solution, the liberated iodine should
be carefully bleached by thiosulphate. Any excess of thiosul-
phate in reasonable amount does not seem to be detrimental,
but in practice the starch iodide color is clearly bleached, and
no more then added. Carbonates prevent a definite indication
of the neutral point by thiosulphate and starch iodide, there-
fore the barium chloride (about 5") should be added before
this point in the process. The mixture of iodide and iodate
is not added to the solution to be analyzed until after it is
made acidic, for the reason that when the neutral point is
approached by the addition of hydrochloric acid, the starch
iodide is thrown out locally by the acid, and the small amount
of sodium borate remaining undecomposed does not again
bleach the coloration produced, thus obscuring the neutral
point which must be obtained before titrating for borie acid.

The solution after the bleaching of the iodine by thiosul-
phate is colorless and contains only starch, neutral chloride,
potassium tetrathionate, iodide and iodate, and all the boric
acid present in uncombined condition. The carbonate lies out
of the sphere of action in insoluble form as barium carbonate.
A few drops of the indicator, phenolphthalein, are now added,
and the alkaline solution run in until a strong red coloration is
produced. A pinch of mannite is then added, which bleaches.
the phenolphthalein coloration, and the alkali solution again

L. C. Jones—Estimation of Borie Acid. 151

run in to a faint indication, which if permanent on the addi-
tion of more mannite, may be taken as the reading point.
About 1-2 grm of mannite are necessary for a determination.
The boro-mannite compound is sufficiently acidic to liberate
iodine abundantly, but it appears to be a time reaction, and at
- the end of six hours only about 95 per cent of the theoretical
amount (considering B,O, as a bivalent acid) has been thrown
out. The combination of boric acid and mannite liberates in
the presence of iodide and iodate immediately only about one-
half the iodine required on the theory that b,O, under these
conditions acts as a bivalent acid, or with the neutralizing
power of metaboric acid, HOBO. When no mannite is present
phenolphthaline gives an alkaline indication when only about
one-half the amount of alkali theoretically necessary to form
the metaborate, NaOBO, has beenadded. Obviously, then, the
starch iodide coloration will not appear at all on the addition
of mannite, if about one-half the free boric acid is first neu-
tralized by the solution of alkali, and the remainder of the
alkali immediately added to complete neutralization. The
point at which the danger of the appearance of the iodide
coloration on the addition of mannite has been passed, is
roughly indicated before the mannite has been added by the
appearance of the strong alkaline indication of phenolphthalein.
This indicator would not need to be added at all, if the boro-
mannite compound quickly and regularly liberated iodine from
the iodide and iodate. The fact, however, that this compound
of boric acid and mannite—as-has been ascertained by experi-
ment——liberates, on standing twelve hours, about 99 per cent
of the theoretical amount of iodine, places the strength of this
acid above that of citric or tartaric acid as investigated by
_ Furry.* With phenolphthalein, however, the end reaction is
sharp and the small amount of carbonate present in the stand-
ard solution of alkali is precipitated by the barium chloride
already in the solution. The calculation must therefore be
based on the amount of free hydroxide in the standard solution
of alkali used, according to the following representation :

B,O,+2Na0H=2Na0BO + H,O.

The best results and the most definite indications are
obtained in cold solution of a volume not greater that
50™*. This fact accords with the observations of Magnaninit
that the relative electrical conductivity of the boro-mannite
solution is decreased by dilution and elevation of the
temperature. When silicates are present in solution, the

* Am. Chem. Jour., vi, 341.
+ Gaz. Chim., xx, 428, and xxi, 134,

152 L. C. Jones—kstimation of Boric Acid.

silicondioxide is liberated by the excess of hydrochloric
acid, and this oxide, whether in hydrous or anhydrous condi-
tion, neither affects the indication with iodine nor phenol-
phthalein, nor does it form with mannite a compound of
acidic properties. Ammonium salts interfere with the indi-
cation given by phenolphthalein and may be removed by boil-
ing with potassium hydroxide in excess, or an indicator used
not affected by them.

To test the action of .fluorides in the process, several experi-

ae n
ments were made in which hydrofluoric acid (10° of =

solution) was introduced into the solution containing salts of
sodium, free hydrochloric and boric acids. Barium chloride
was then added and the analysis for boric acid completed in
the usual way without the aceuracy of the results being in any
way interfered with by the presence of hydrofluoric acid.

The following table contains the results of a series of anal-
yses in which the boric acid was first drawn into an excess of
sodium hydroxide, then estimated according to the method
described.

The standard solutions of boric acid used contained I, 7-153
grm. and II, 7-706 grm. per liter. The solution of free sodium
hydroxide was 0°21427 normal.

TABLE I.

B.0; Sol. NaOH Sol. B,O; B.03 Errors on
taken. required. taken. found. B.Os.
em? em? erm, erm. grm.

(ih) is, 20295 21°02 eS 7k 0°1577 +0 0006

= (2) 20°68 19°65 071479 0:1474 —0°0005
(3) 20°73 19°63 0°1483 0°1473 —0°0010

( (4) 23°05 93°71 0°1776 OL Lea +0°0001

| (5) 23-10 23°80 0°1780 01783 + 0°00038
= (G)y"7 22°76 25°35 0°1754 0°1750 — 0'0004
(ay 24-08 Ps OATS 0°1855 0°1857 +0°0002
(8) 22-00 22°50 0°1695 0°1686 —0°0009

| (9) 20°78 21°28 0°1601 0°1595 — 0°0006

Practical tests of the method were made upon specimens of
crude calcium borate and colemanite.*

The finely-ground minerals were dissolved in hydrochloric
acid and the analyses proceeded with as above described.

* These specimens were kindly furnished by Dr. C. A. Crampton, of Washing-
ton, whom | desire to thank for this courtesy.

L. ©. Jones—Estimation of Boric Acid. 153

Analysis of crude borate of lime.

TABLE II.
Ca borate BQ; found. B.03;
taken. grm. grm. in %.
(1) 0°4016 0°2289 56°99
(2) 0-4044 0°2302 56°92
(3) 0°4000 0°2285 LH ea |
Analysis of Colemanite.
TABLE III.
Mineral taken. BO; found. _ % BOs.
erm. erm.
(1) 0°4034 0°2064 51°15
(2) 0°4070 0°2069 50°80 Average
(3) 0°6004 0°3054 50°86 50-994
(4) 0°6006 0°3056 50°89: ‘
(5) 0°5059 0°2592 91°24
(6) 0°5092 0°2592 50°89

An analysis for boric acid by this process can be completed
in five minutes and the results are obviously accurate within
the limits of ordinary analysis.

The usually interfering substances, fluorine, silica, and car-
bon dioxide, have no detrimental influence on the results of
this process.

154 Scientific Intelligence.

SCIENTIFIC INTELLIGENCE.

I. CHEMISTRY AND PHYSICS.

; 1. On the Molecular Masses of Gases.—It has been shown by

D. BertHeELot that in consequence of the difference between the
coeflicients of expansion and compressibility, the molecular masses.
of gases are proportional, not to the actual density, but to the
product of this value into the molecular volume. The molecular

volume of a gas is given by the formula v, = es [1+ a(p—p,)|=

yh (1+ap—ap,); in which v is the actual volume under an infin-
itely small pressure p, p, is the normal pressure, and @ is a coefli-
cient, constant, according to Regnault between 1 and 6 atmospheres
and according to Amagat, below one atmosphere. When 7p, is one
atmosphere and: p is indefinitely small, the ratio between the
molecular volumes of two gases is v,/v,=(l—a)/(1-a@). If
their densities be dand d’ their molecular masses are propor-
tional to (1—a)d and (1—a@’) ad’. In the table below, the value of
a at 0°, the molecular volume v,, at 0° and one atmosphere, the
specific gravity d, oxygen being taken as a standard, the molecu-
lar mass M when O = 32 and the atomic mass w when O = 16, are
given as calculated :

H N CO O
2s See — 0°00046 +0°00038 +0:00046 +40:00076
Vgc ieeeere 1°00046 0°99962 0-99954 0°99924
Os eee 0°062865 0°87508 0°87495 1°0
Mi ae 2°01472 28°01382 28°0068 32
TIME 9 oes = 1:0074 14°007 12-007 16

The author compares the atomic masses of hydrogen, nitrogen,
carbon and oxygen deduced as above from the molecular vol-
umes and relative densities, with the best determinations of these
values by chemical methods, and shows that they are almost
identical.

O Jal N C

Physical method, 16 1°:0074 14:007 12°007
Chemical method, 16 1:0023—1°0075 14:012 12°001—12-005

It would appear therefore that not only do the values of the
atomic masses obtained from relative density and compressibility
confirm the results obtained by chemical methods, but that in
some cases they may be taken as the more reliable of the two.—
C. &., exxvi, 954-956, 1030-1033, March, April, 1898. «G. F. B.

2. On the Preparation of Metals and Alloys by means of
Aluminum. — The commercial production of aluminum on a
large scale has rendered it possible to make use of this metal for
the reduction of the oxides and chlorides of other metals.

Chemistry and Physics. 155

Experiments by Gotpscumipt have greatly improved the earlier
processes, by showing that it is not necessary to heat the crucible
containing the mixture, from the outside, as in a furnace. It is
quite sufficient to inflame this mixture at a single point, from -
which the action spreads more or less quickly throughout the
mass. The ignition may be effected by means of a piece of mag-
nesium wire inserted in the mass and lighted by a match. For
mixtures of higher igniting points, a primer made by mixing
aluminum powder with a substance easily giving up oxygen, such
as lead or manganese peroxide, lead or copper oxide, may be
placed on the mixture and this inflamed by the magnesium wire.
Since the temperature of the reaction is intense, it is best to line
the crucible with a refractory material such as magnesia or
alumina, or in some cases the oxide of the metal that is to be
reduced. The process may be modified by igniting first a small
quantity of the mixture in the crucible, and then adding addi-
tional portions from time to time. The metals thus obtained are
free from carbon and from carbide, and from aluminum also if the
oxide is present in excess. In the case of chromium, the tem-
perature reached was found approximately to be 3000°. The
fused chromium thus prepared is permanent even in the air of the
laboratory, behaving like a noble metal. Fused manganese
resembles bismuth, showing surface colors. By this process,
alloys of iron with 20 to 25 per cent of boron and with 40 per
cent of titanium have been prepared; and also copper containing
10 per cent of chromium. An alloy of lead and barium decom-
poses water. Sulphides may be reduced in this way and ata
lower temperature. The resulting slag when oxides are used is
fused alumina, so hard as scarcely to be scratched by the diamond.
In preparing chromium, minute transparent red crystals, resem-
bling rubies, are found in it.—Liebig’s Annalen, ccci, 19, June,
1898. G. F. B.

3. On the Electrolytic Preparation of Beryllium.—It has been
shown by Lexeat that beryllium fluoride, while a non-conductor
of electricity when fused, becomes a conductor and an electrolyte
when mixed with an alkali fluoride. Two beryllium-sodium
fluorides, BeF(NaF), and Bek’. NaF, are available for this pur-
pose. They are most easily prepared by dissolving beryllium
hydrate and sodium carbonate, in the required proportions, in
hydrogen fluoride and evaporating. The first of these double
chlorides fuses at about 350° and the other at about redness. The
salt to be reduced is melted in a nickel crucible, acting as the
cathode, a rod of graphitic carbon serving as the anode. As
soon as the current passes, the external source of heat is with-
drawn, the temperature, which should not be allowed to rise
above a low red heat, being thereafter maintained electrically.
A current of 6 to 7 amperes is required, at an electromotive
force of 30 to 40 volts; the metal separating in the crystalline
form. If the electrolysis is conducted in a carbon crucible,
which contains the other metal in the fused state, and which

156 Scientific Intelligence.

serves as the anode, alloys of beryllium may be obtained.—C. R.,
exxvi, 744-746, March, 1898. G. B.. B:

4. On Colloidal Silver.—Some experiments have been made by
LotTTERMOSER and von MEYER on the action of various reagents
on colloidal silver. They find that acids precipitate it from solu-
tion in the “molecular” state, time and dilution being of great
importance. To study the influence of time on the change from
the colloidal to the molecular condition, a known amount of acid
was added to the colloidal solution and the time was observed
which was required for the complete change. By placing a drop
of the solution on a glass plate resting on white paper, the end of
the reaction is easily detected; the time being noted as soon as
the separation of fine silver particles can be readily seen. The
results of the experiments are given in a table. They show that
the less the amount of acid used the longer the time required for
the reaction. Moreover, the amount of acid required to produce
the change in a given time appears to be inversely as the concen-
tration of the solution. As to the precipitation of colloidal silver
by salts of the heavy metals, experiment shows that where reduci-
ble metallic chlorides are used there is formed silver chloride and
a lower chloride of the metal; a portion of the silver, however,
being always precipitated in the insoluble form. If the solution
of colloidal silver and of the metallic chloride are both very
dilute, it frequently happens that neither silver chloride nor the
lower metallic chloride is precipitated, both remaining in solution
in the colloidal form. This behavior is particularly shown by
mercuric chloride. Neither the chloride, bromide or iodide of
colloidal silver appears to have any application in photography,
though the iodide is of interest therapeutically.—J. pr. Ch.,
lvi, 241, 1897; lvii, 540, July, 1898. G. F. B.

5. On Silver peroxide and peroxynitrate.—In continuing his
researches on the higher oxide of silver and its compounds
Muuper has observed that when the substance Ag,NO,, is de-
prived of an atom of oxygen and the residue is extracted with
water, silver nitrate goes into solution and silver peroxide Ag,O,
remains. Since the silver nitrate can be extracted in a short
time, the nitrate Ag,NO,, may be regarded as having the compo-
sition (Ag,O,),. AgNO,. When boiled with water the peroxy-
nitrate is decomposed, oxygen is set free and silver nitrate goes
into solution, leaving a residue of silver peroxide. This suggests
a convenient method for preparing the peroxide. This higher
oxide dissolves in concentrated sulphuric and in concentrated nitric
acid, giving brown solutions. The presence of water accelerates
the decomposition of these solutions. The peroxide is not solu-
ble in acetic acid though when diluted silver acetate and oxygen
are formed. A method for determining the excess of oxygen in
the peroxynitrate is described, based on the decomposition of this
salt by water. When kept for some time the peroxynitrate un-
dergoes slow spontaneous decomposition even over sulphuric acid.
So slow is this change, however, that the author estimates that it

Chemistry and Physics. 157

would require about 13 years to eliminate the two atoms of oxy-
gen in excess.— Rev. Trav. Chim., xvii, 129-176, 1898; J. Chem.
Soc., Ixxiv, 11, 516, Nov. 1898. G. F. B.

6. On Neodymium.—According to BoupovuaRD, neodymium can
be isolated by allowing a solution of the pure sulphates of the
yttrium metals to remain in contact with potassium sulphate in
excess for at least 24 hours, then decomposing the insoluble
double sulphate thus produced, with sodium hydrate, dissolving
the oxide in nitric acid and precipitating with oxalic acid. The
substance thus prepared afforded the atomic mass 143, agreeing
well with 142-7 ordinarily assigned to this metal. The oxide is
greenish, while the oxalate and anhydrous sulphate are slightly
rose-colored. The crystallized sulphate, also rose-colored, while
less soluble than the anhydrous sulphate, is more soluble in cold
water than in hot. A solution of the sulphate gives an absorption
spectrum having a shadow from 591°5 to 584, an intense band from
_ 584 to 572, another from 523 to 519, a feeble band from 512 to
508, and faint bands at 480 and 470, due probably to traces of
praseodymium. The double sulphate of potassium and neody-
mium is more soluble than the corresponding compound of pra-
seodymium.— C. #., exxvi, 9€0—-901, March, 1898. G. F. B.

7. Legons de Chimie Physique professees ad 0 Université de
Berlin ; par J. H. Van’t Horr, Professeur ordinaire a ?Univer-
sité, etc. Ouvrage traduit de l’allemande par M. Corvisy. Pre-
miére Partie, La Dynamique Chimique. 8vo, pp. 263. Paris,
1898 (A. Hermann).—The lectures delivered by Professor Van’t
Hoff during 1897 were upon ‘Selected subjects in Physical Chem-
istry” and were divided into three parts, the first comprising
Chemical Dynamics, the second Chemical Statics, and the third
Relations between the Properties of Substances and their Compo-
sition. The present volume contains the first part only; and like
all the productions of this eminent authority, is full of the most
valuable material admirably classified and clearly treated. The
subject is divided into two principal sections, the first on Chemi-
eal Equilibrium and the second on Speed of Reaction; each of
these being subdivided into two parts, the former section into
Chemical Equilibrium in its Physical and especially its Thermo-
dynamic relations and Chemical Equilibrium in its Mechanical
and Molecular relations; and the latter section into Theoretical
laws of reaction speeds and Experimental results of Reaction
speeds. The translator seems to have done his work with care
and thus to have made available this excellent presentation of the
subject to a much larger class of students. G. F, B.

8. Outlines of Industrial Chemistry ; a Text-book for Students.
By Frank Haut Tuorp, Ph.D. 8vo, pp. xx, 541. New York,
1898 (The Macmillan Company).—Dr. Thorp has given us in
this book a compendium of commercial chemical processes which
cannot fail to be of use to the technicalstudent. The first part is
devoted to Inorganic industries such as relate to fuels, waters,
the soda industry, the chlorine industry, fertilizers, cements, glass

158 Scientific Intelligence.

and ceramics, pigments, etc. The second part treats of Organic
industries, connected with the destructive distillation of wood,
bones and coal, with mineral oils, with vegetable and animal oils
and soaps, with the sugars and fermentation, with explosives,
textiles, etc. The volume is well printed and illustrated. G. F. B.

9. Charge of Electricity carried by the ions produced by
Réntgen Rays.—Mr. C. T. R. Wilson (Phil. Trans. A., 1897, p.
265\ has discovered that Réntgen rays can produce a cloudy con-
densation in dust-free air when the latter is subjected to a sudden
expansion. This condensation cannot be produced by expan-
sion without the aid of the rays. Prof. J. J. Tuomson has
made use of this phenomenon to determine the value of the prod-
uct 2 €v where 7 is the number of ions in unit volume of the gas,
e the charge in an ion, v the mean velocity of the positive and
negative ions under a definite electromotive force. The size of the

2
drops was determined from the expression v = ce in which g

is the acceleration of gravitation ; a, the radius of the drop
around the ions as nuclei, p» the coefticient of viscosity of the gas
through which the drops fell. The velocity was determined by
observing the time the top layer of the cloud took to fall a given
distance.

The mean value of the charge on the ion was found to be
6°5x<107"*. The experiments seemed to show that the charge on
the ions in hydrogen was the same as in air. From the laws of
electrolysis, if e is the charge on the hydrogen ion in electrostatic
units, V the number of molecules in 1 cub. centim. at standard
temperature and pressure,

NES N29 ee If e is

taken as 655210 ge

I= QOPSCINOS:
When XN deduced from experiments on viscosity of air is
21x10". The agreement between the value of WV got by the
kinetic theory of gases by viscosity experiments and the value
obtained by Professor Thomson’s experiments, leads him to believe
that the theory is consistent with the value he has obtained for e,
being equal to or of the same order, as the charge carried by the
hydrogen ion in electrolysis.— Phil. Mag., Dec. 1898, pp. 528—
545, Jacks

10. Use of the Coherer in Measuring Electric Waves.—It is
still a disputed question whether the coherer can be used in the
investigation of stationary electric waves. Professor Murani, of
Milan, has published an investigation which seems to show that
the coherer is too sensitive to such waves, and that having
responded to the first electric impulse it is no longer sensitive to
* succeeding ones. One thus obtains a more or less constant deflec-
tion through a galvanometer in the coherer circuit without any
evidence of nodes or ventral segments. O. BEHRENDSEN believes
that better results can be obtained by using a comparatively in-
sensitive coherer, which will only respond “to the integrated im-

Chemistry and Physics. 159

pulses. He makes such acoherer out of comparatively fine carbon
powder. By means of such a coherer he obtained evidences
of interference phenomena.. With a sensitive coherer filled with
nickel filings he measured the dimension of electric waves
produced by interference in passing through a metallic grating.
— Wied. Ann., No. 18, 1898, pp. 1024-1029. a Ts
11. Color Blindness and the Réntgen Rays.—E. Dorn has in-
vestigated the sensitiveness of the eyes of a totally color-blind
person to the Rontgen rays. He believes that the light seen by
the color-blind person when the eye is exposed to the rays is not
due to a fluorescence excited in the retina by these rays, but is due
to a peculiar aptitude of the eyes to perceive these rays. The
light excited by the rays is greatest at the periphery of the eye;
this indicates that the rods of the eye are sensitive to the Ront-
gen rays. The author then sought to answer the question
whether the cones are also sensitive. He discovered that these
are also sensitive and no central blind spot insensitive to these
rays could be found in the totally color-blind eye. His experi-
ments he believes can be used to disprove the hypothesis that
total color-blindness can be ascribed either to failure or the apti-
tude of the cones of the eye.— Wied. Ann., No. 13, 1898, pp. 1170-
1176. diage:
12. Regenerating Vacuum Tubes; by W. Roxins. (Com-
municated.)—In an interesting article in the January number of
this Journal on absorption of gases in a high vacuum, Prof. C. C.
Hutchins states that after a short period of work the vacuum in
a Roéntgen tube becomes too high. He then considers the causes
of the rise and discusses the various methods for increasing the
pressure, recommending oxygen as the most suitable gas because
its absorption is slow. The use of oxygen is not new; tubes with
adjustable oxygen vacuums were made soon after Rontgen’s
discovery was announced and were described in a series of notes
on Rontgen light in the Electrical Review in 1897-8, and are illus-
trated in Kirmayer and Oelling’s catalogue. In this connection it
may be well to state that regenerative tubes with hydrogen vacuums
have been recently recommended by Villard, who attaches a closed
platinum tube which is heated in a Bunsen flame when it is de-
sired to increase the pressure in the Réntgen tube. <A regener-
ating tube of this kind was described in 1897 in the above-men-
tioned notes, which was superior to the more recent form, because
the hollow platinum tube was used as a target for the impact of
the cathode discharge. As this hollow target could be cooled, a
large amperage could be used. ‘To increase the pressure it is
only necessary to put a drop of liquid hydrocarbon into the hol-
low target and heat it with the cathode discharge, or illuminating
gas can be used, retaining it temporarily in place with a bit of
cork, 3
In some of my experiments I tried a plan used by Professor
Trowbridge soon after Réntgen’s discovery was announced, but
now lost sight of, which consisted in cooling the discharge tube

160 Sceventifie Intellagence.

in oil. When this method is applied to a hollow target tube the
fresh, cool oil must be made to constantly strike against the back
of the target, otherwise this will get réd hot and drive back the
oil, the hydrogen passing through the platinum. In this way I
have had the vacuum change from fifteen inches to one-sixteenth
of an inch equivalent air-spark in a short time.

What those of us who are trying to use Roéntgen’s discovery
in medical diagnosis want, is not rediscoveries, but to have placed
within our reach apparatus which will convert into Réntgen light
an appreciable amount of the energy put into it. At present we
may pour in at one end energy at the rate of ten amperes at one
hundred volts, and when we come to take it out at the other end
as Rontgen radiation there will not be enough with good defini-
tion to clearly see through the abdomen. If some broad commer-
cial use could be found for Réntgen’s discovery, then the great
inventors would attack the problem. At present a tallow candle,
notwithstanding its thermic waste, is as a converter of energy
into light a shining success compared with the best available
Réntgen apparatus. The only notable advance which has been
made in practice is due to M. Tesla, but even he seems unable to
give much time to this matter; asa result, neither he nor anyone
knows when his apparatus will be commercially available. Even
when we get it the results will probably be far below the possi-
bilities of the method as shown by the hints dropped from time
to time by prominent physicists. Take, for example, Professor
Trowbridge’s statement that he had made a photograph of the
bones of a hand in one-millionth of a second. With the best
available apparatus the exposure must be from sixty million to
three hundred million times as long. Part of the advance which
will be made must consist in devising simple means of making
the exciting surges harmonics of the rate of vibration we wish to
produce in the radiant energy. All my experiments point in this
direction, and [ find it practical to get a better result with a
small properly-tuned generator than with the largest size improp-
erly adjusted, but I know too little of physics to make any impor-
tant advance in this matter. .

II. Grouocy AND NATURAL HISTORY.

1. The Age of the Harth.—The January number of the Philo-
sophical Magazine contains a notable article by Lorp Ketvin,
on the age of the earth as an abode fitted for life. This is an
extension of the Victoria Institute Annual Address for 1897. The
opening pages give an interesting account of the early attitude
of the geologist to the subject in demanding almost unlimited
time for the geological changes the earth has gone through and
for the development of life. Then follows a summary of the
arguments by which the author showed (1862 to 1869) the strict
limitations of the possible age of earth, viewed as an abode for
life, and arrived at the conclusion that the earth’s consolidation

Geology and Natural History. 161

could not have taken place less than 20 million, nor more than
400 million years ago. With respect to this he adds:

“ During the 35 years which have passed since I gave this wide-
ranged estimate, experimental investigation has supplied much of
the knowledge then wanting regarding the thermal properties of
rocks to form a closer estimate of the time which has passed since
the consolidation of the earth, and we have now good reason for
judging that it was more than 20 and less than 40 million years
ago; and probably much nearer 20 than 40.”

Remarking upon the conclusion reached by Mr. Clarence King
(this Journal, Jan., 1893), based largely upon the experimental
data supplied by Dr. Carl Barus, in the laboratory of the U. S.
Geological Survey, that “‘ we have no warrant for extending the
earth’s age beyond 24 million of years,” he adds that the results
of his own recent calculations lead to an estimate not differing
much from this.

Going on to discuss the course of events following the first
solidification of the earth’s surface, the author says :

“$21. I have given strong reasons* for believing that émmedi-
_ ately before solidification at the surface, the interior was solid
close up to the surface: except comparatively small portions of
lava or melted rock among the solid masses of denser solid rock
which had sunk through the liquid, and possibly a somewhat
large space around the center occupied by platinum, gold, silver,
lead, copper, iron, and other dense metals, still remaining liquid
under very high pressure.”

““§ 22. I wish now to speak to you of depths below the great
surface of liquid lava bounding the earth before consolidation ;
and of mountain heights and ocean depths formed probably a
few years after a first emergence of solid rock from the liquid
surface (see § 24, below), which must have been quickly followed
by complete consolidation all round the globe.” .. .

““§ 23. To prepare for considering consolidation at the surface
let us go back to a time (probably not more than twenty years
earlier as we shall presently see—§ 24) when the solid nucleus was
covered with liquid lava to a depth of several kilometers; to fix
our ideas let us say 40 kilometers (or 4 million centimeters). At
this depth in lava, if of specific gravity 2°5, the hydrostatic pres-
sure is 10 tons weight (10 million grammes) per square centime-
ter, or ten thousand atmospheres approximately. According to
the laboratory experiments of Clarence King and Carl Barust on
diabase, and the thermodynamic theory{ of my brother, the late

*“On the Secular Cooling of the Earth,” vol. iii, Math. and Phys. Papers,
§§ 19-33.

+ Phil. Mag., 1893, first half-year, p. 306.

t{ Trans. Roy. Soc, Edinburgh, Jan. 2, 1849; Cambridge and Dublin Mathe-
matical Journal, Nov., 1850. Reprinted in Math. and Phys. Papers (Kelvin), vol.
i, p. 156.

Am. Jour. Sct.—Fourta Series, Vou. VII, No. 38.—FEsrRuARyY, 1899.
11

162 Scientific L ntelligence.

Professor James Thomson, the melting temperature of diabase is
1170° C. at ordinary atmospheric pressure, and would be 1420°
under the pressure of ten thousand atmospheres, if the rise of
temperature with pressure followed the law of simple proportion
up to so high a pressure.”

“$24. The temperature of our 40 kilometers deep lava ocean
of melted diabase may therefore be taken as but little less than
1420° from surface to bottom. Its surface would radiate heat.
out into space at some such rate as two (gramme-water) thermal
units Centigrade per square centimeter per second.* Thus, in a
year (315 million seconds) 63 million thermal units would be lost
per square centimeter from the surface. This is, according to
Carl Barus, very nearly equal to the latent heat of fusion aban-
doned by a million cubic centimeters of melted diabase in solidi-
fying into the glassy condition (pitch-stone) which is assumed
when the freezing takes place in the course of a few minutes.
But, as found by Sir James Hall in his Edinburgh experimentst
of 100 years ago, when more than a few minutes is taken for the
freezing, the solid formed is not a glass but a heterogeneous crys-
talline solid of rough fracture; and if a few hours or days, or any
longer time, is taken, the solid formed has the well known rough .
crystalline structure of basaltic rocks found in all parts of the
world. Now Carl Barus finds that basaltic diabase is 14 per
cent denser than melted diabase, and 10 per cent denser than the
glass produced by quick freezing of the liquid. He gives no
data, nor do Riicker and Roberts-Austen, who have also experi-
mented on the thermodynamic properties of melted basalt, give
any data, as to the latent heat evolved in the consolidation of
liquid lava into rock of basaltic quality. Guessing it as three
times the latent heat of fusion of the diabase pitch-stone, I esti-
mate a million cubic centimeters of liquid frozen per square cen-
timeter per centimeter per three years. This would diminish the
depth of the liquid at the rate of a million centimeters per three
years, or 40 kilometers in twelve years.”

‘$25. Let us now consider in what manner this diminution of
depth of the lava ocean must have proceeded, by the freezing
of portions of it; all having been at temperatures very little
below the assumed 1420° melting temperature of the bottom,
when the depth was 40 kilometers. The loss of heat from the
white-hot surface (temperatures from 1420° to perhaps 1380° in
different parts) at our assumed rate of two (gramme-water Centi-
grade) thermal units per sq. cm. per sec. produces very rapid
cooling of the liquid within a few centimeters of the surface

* This ig a very rough estimate which I have formed from a consideration of
J. T. Bottomley’s accurate determinations in absolute measure of thermal radia-
tion at temperatures up to 920° C. from platinum wire and from polished and
blackened surfaces of various kinds in receivers of air-pumps exhausted down to
one tenth-millionth of the atmospheric pressure. Phil. Trans. Roy. Soc., 1887
and 1893.

+ Trans. Roy. Soc. Edinburgh.

Geology and Natural History. 165

(thermal capacity ‘36 per gramme, according to Barus) and in
consequence great downward rushes of this cooled liquid, and
upwards of hot liquid, spreading out horizontally in all direc-
tions when it reaches the surface. When the sinking liquid gets
within perhaps 20 or 10 or 5 kilometers of the bottom, its tem-
perature* becomes the freezing-point as raised by the increased
pressure; or, perhaps more correctly stated, a’ temperature at
. which some of its ingredients crystallize out of it. Hence, be-

ginning a few kilometers above the bottom, we have a snow
shower of solidified lava or ef crystalline flakes, or prisms, or
granules of feldspar, mica, hornblende, quartz, and other ingre-
dients: each little crystal gaining mass and falling somewhat
faster than the descending liquid around it, till it reaches the bot-
tom. This process goes on until, by the heaping of the granules
and crystals on the bottom, our lava ocean becomes silted up to
the surface.”

“ Probable Origin of Granite.—§ 26. Upon the suppositions we
have hitherto made, we have, at the stage now reached, all round
the earth at the same time a red hot or white hot surface of solid

ranules or crystals with interstices filled by the mother liquor
still liquid, but ready to freeze witb the slightest cooling. The
thermal conductivity of this heterogeneous mass, even before the
freezing of the liquid part, is probably nearly the same as that of
ordinary solid granite or basalt at a red heat, which is almost
certainly+ somewhat less than the thermal conductivity of igneous
rocks at ordinary temperatures. If you wish to see for yourselves
how quickly it would cool when wholly solidified, take a large
macadamizing stone, and heat it red hot in an ordinary coal fire.
Take it out with a pair of tongs and leave it on the hearth, or on
a stone slab at a distance from the fire, and you will see that in a
minute or two, or perhaps in less than a minute, it cools to below
red heat.”

Ǥ 27. Half an hourt after solidification reached up to the sur-
face in any part of the earth, the mother liquor among the gran-
ules must have frozen to a depth of several centimeters below the
surface and must have cemented together the granules and crys-
tals, and so formed a crust of primeval granite, comparatively
cool at its upper surface, and red hot to white hot, but still all
solid, a little distance down; becoming thicker and thicker very
rapidly at first; and after a few weeks certainly cold enough at
its outer surface to be touched by the hand.”

* The temperature of the sinking liquid rock rises in virtue of the increasing
pressure: but much less than does the freezing point of the liquid or of some of
its ingredients. (See Kelvin, Math. and Phys. Papers, vol. iii, pp. 69, 70.)

+ Proc. Roy. Soc., May 30, 1895.

t Witness the rapid cooling of lava running red hot or white hot from a vol-
cano, and after a few days or weeks presenting a black hard crust strong enough
and cool enough to be walked over with impunity.

164 - Serentific Intelligence.

“‘ Probable Origin of Basaltic Rock.*—§ 28. We have hitherto
left, without much consideration, the mother liquor among the
cr ystalline granules at all depths ‘below the bottom of our shoal-
ing lava ocean. It was probably this interstitial mother liquor
that was destined to form the basaltic rock of future geological
time. Whatever be the shapes and sizes of the solid grauules
when first falling to the bottom, they must have lain in loose
heaps with a somewhat large proportion of space occupied by
liquid among them. But, at considerable distances down in the
heap, the weight of the ‘superincumbent granules must tend to
crush corners and edges into fine powder. If the snow shower
had taken place in air we may feel pretty sure (even with the
slight knowledge which we have of the hardness of the crystals
of feldspar, mica and hornblende, and of the solid granules of
quartz) that, at a depth of 10 kilometers, enough of matter from
the corners and edges of the granules of different kinds, would
have been crushed into powder of various degrees of fineness, to
leave an exceedingly small proportionate volume of air in the in-
terstices between the solid fragments. But in reality the effect-
ive weight of each solid particle, buoyed as it was by hydrostatic
-pressure of a liquid less dense than itself by not more than 20 or
15 or 10 per cent, cannot have been more than from about one-
fifth to one-tenth of its weight in air, and therefore the same de-
gree of crushing effect as would have been experienced at 10
kilometers with air in the interstices, must have been experienced
only at depths of from 50 to 100 kilometers below the level of
the lava ocean.”

“«§ 29. A result of this tremendous crushing together of the
solid granules must have been to press out the liquid from among
them, as water from a sponge, and cause it to pass upwards
through the less and less closely packed heaps of solid particles,
and out into the lava ocean above the heap. But, on account of

* NOTE BY THE FpitoR: An Addendum at the close of the paper quotes the
following determination of melting points by Prof. Roberts- Austen:

Melting-point. Error.
elds pan, se. Wee ee h20"C. +30°
Hornbplende: 2 22). e saan about 1400°
WiiGaet 7 (Seer ae ee 1440° + 30°
Quartz: 2s Me eae 1175° +15°
Basalt {tis ic hexanal eaten about 880°

The author adds:

‘These results are in conformity with what I have said in §§ 26-28 on the
probable origin of granite and basalt, as they show that basalt melts at a much
lower temperature than feldspar, hornblende, mica, or quartz, the crystalline in-
gredients of granite. In the electrolytic process for producing aluminium, now
practised by the British Aluminium Company at their Foyers works, alumina, of
which the melting-point is certainly above 1700° C. or 1800° C, is dissolved in a
bath of melted cryolite at a temperature of about 800° C. So we may imagine
melted basalt to be a solvent for feldspar, hornblende, mica, and quartz at tem-
peratures much below their own separate melting-points; and we can understand
how the basaltic rocks of the earth may have resulted from the solidification of
the mother liquor from which the crystalline ingredients of granite have been
deposited.”

Geology and Natural History. 165

the great resistance against the liquid permeating upwards 30 or
40 kilometers through interstices among the solid granules, this
process must have gone on somewhat slowly ; and, during all the
time of the shoaling of the larva ocean, there may have been a
considerable proportion of the whole volume occupied by the
mother liquor among the solid granules, down to even as low as
50 or 100 kilometers below the top of the heap, or bottom of the
ocean, at each instant. When consolidation reached the surface,
the oozing upwards of the mother liquor must have been still
going on to some degree. Thus, probably for a few years after
the first consolidation at the surface not probably for as long as
one hundred years, the settlement of the solid structure by mere
mechanical crushing of the corners and edges of solid granules,
may have continued to cause the oozing upwards of mother liquor
to the surface through cracks in the first formed granite crust and
through fresh cracks in basaltic crust subsequently formed
above it.”

After a further discussion of the probable origin of continents
and ocean depths, and also of the earth’s atmosphere, the author
concludes as follows:

“43, Whatever may have been the true history of our atmo-
sphere it seems certain that if sunlight was ready, the earth was
ready, both for vegetable and animal life, if not within a century,
at all events within a few hundred centuries after the rocky con-
solidation of its surface. But was the sun ready? The well
founded dynamical theory of the sun’s heat carefully worked out
and discussed by Helmholtz, Newcomb, and myself,* says NO if
the consolidation of the earth took place as long ago as 50 mil-
lion years; the solid earth must in that case have waited 20 or 50
million years for the sun to be anything nearly as warm as he is
at present. If the consolidation of the earth was finished 20 or
25 million years ago, the sun was probably ready,—though prob-
ably not then quite so warm as at present, yet warm enough to
Support some kind of vegetable and animal life on the earth.”

“$44. My task has been rigorously confined to what, bumanly
speaking, we may call the fortuitous concourse of atoms, in the
preparation of the earth as an abode fitted for life, except in so
far as I have referred to vegetation, as possibly having been con-
cerned in the preparation of an atmosphere suitable for animal
life as we now have it. Mathematics and dynamics fail us when
we contemplate the earth, fitted for life but lifeless, and try to
imagine the commencement of life upon it. This certainly did
not take place by any action of chemistry, or electricity, or crys-
talline grouping of molecules under the influence of force, or by
any possible kind of fortuitous concourse of atoms. We must
pause, face to face with the mystery and miracle of the creation
of living creatures.” .

* See “Popular Lectures and Addresses,” vol. i, pp. 376-429, particularly page
397.

166 Scientific Intelligence.

2. Recent Publications of the U. 8. Geological Survey :*
ANNUAL Reports.—The five parts (making 6 volumes) of the 78th
Annual Report are now complete. All have been noticed except-
ing the following papers included in parts II and IL:

“The Triassic Formations of Connecticut, by W. M. Davis. Pt.
II, pp. 1 to 192, 20 pls. including folded map and sections.

Geology of the Hdwards Plateau and Rio Grande Plain adja-
cent to Austin and San Antonio, Texas, with reference to the
occurrence of underground waters, by R. T. Hill and T. W.
Vaughan. Pt. H, pp. 198-321, pls. 31.

A Table of the North American Tertiary Horizons, correlated
with one another and with those of Western Hyurope, with annota-
tions, by W. H. Dall... Pt. IT, pp. 323 to 348.

Glaciers of Mount Rainier, by I. C. Russell, with a paper on
The Rocks of Mount Rainier, by G. O. Smith. Pt. II, pp. 349
to 423, pls. 18.

The Age of the Franklin White Limestone of Sussex County,
New Jersey, by J. HK. Wolff and A. H. Brooks. Pt. II, pp. 425 to
457, with geological map.

A Geological Sketch of San Clemente Island, by W. 8. T.
Smith. Pt. II, pp. 459 to 496, pls. 13.

Geology of the Cape Cod District, by N. 8S. Shaler. Pt. IT,
pp. 497 to 593, pls. 8.

[Recent Earth Movement in the Great Lakes Region, by G. K.
Gilbert. Pt. Il, pp. 595 to 647, pl. 1.

Some Coal Kields of Puget Sound, by Bailey Willis. Pt. IIL,
pp. 393 to 436, 17 pls.

Geology and Mineral Resources of the Judith Mountains of
Montana, by W. H. Weed and L. V. Pirsson. Pt. IIT, pp. 439 to
616, 18 pls. including maps. See this Journal, vol. vi, p. 508,
December, 1898.

The Mining Districts of the Idaho Basin and the Boise hidge,
Idaho, by Waldemar Lindgren, with a report on fossil plants by
F. H. Knowlton. Pt. III, pp. 619 to 794, pls. 16.

Preliminary Report on the Mining Industries of the Telluride
Quadrangle, Colorado, by C. W. Purington. Pt. III, pp. 745 to
850. The areal geology of this quadrangle is given by Mr. Cross
in plate cill.

The Nineteenth Annual Report (1897-98) will appear in six
parts (seven volumes). The report of the Director C. D. Walcott
of Part I, pp. 148 with two folded maps, giving an account of the
operations of the Survey to June 30, 1898, is already issued. The
total appropriation for the work of the Survey was $1,033,963.60.
Besides continuing the work already begun along essentially the
same lines, this increased appropriation provided for the Survey of
Forest Reserves, the boundary line between Idaho and Montana
and the extension of work in Alaska. Mineral Resources for

* Issued since March, 1898. For last complete list and references to earlier
ones see this Journal for April, 1898, p. 303; notices published since this date
are referred to under the respective titles.

Geology and Natural History. 167

1897, forming Part VI of the Nineteenth Annual Report, will
appear in two volumes; the first including Metallic Products,
Coal and Coke, the second, Non-Metallic Products excepting
Coal and Coke. Separates are nearly all issued and volumes will
soon appear.

Monoerapus: Fossil Meduse Mon. xxx, by C. D. Walcott,
pp- 201, pls. 47. See this Journal, December, 1898, p. 508.

Buuietins: The Cretaceous Foraminifera of New Jersey, by
R. M. Bagg, Jr. Bull. No. 88, pp. 89, pls. 6.

Some Lava Flows of the Western Slope of the Sierra Nevada,
California, by F. L. Ransome, Bull. 89, pp. 73. See this Journal
for December, 1898, p. 509.

Bibliography and Index of North American Geology, Paleon-
tology, Petrology and Mineralogy, by F. B. Weeks, Bull. No.
149, pp. 152, for 1896; Bull. 156, pp. 130, for 1897.

The Educational Series of Rock Specimens collected and dis-
tributed by the U. 8. Geological Survey, by J. 8. Diller, Bull.
150, pp. 400, pls. 47. See this Journal, January, 1899, p. 74.

The Lower Cretaceous Grypheas of the Texas Region, by R.
T. Hill and T. W. Vaughan, Bull. No. 151, pp. 138, pls. 35. See
this Journal, January, 1899, p. 70.

A Catalogue of Cretaceous and Tertiary Plants of North
America, by F. H. Knowlton, Bull. 152, pp. 247.

A Bibliographic Index of North American Carboniferous
Invertebrates, by Stuart Wheeler, Bull. No. 153, pp. 653.

A Gazetteer of Kansas, by Henry Gannett, Bull. No. 154, pp.
246, with map of the State.

Earthquakes in California in 1896 and 1897, by C. D. Perrine,
Bull. 155, pp. 45.

Water Suprpiy AND IRRIGATION Papers: No. 12, Underground
Waters of a portion of Southeastern Nebraska, by N. H. Darton,
pp. 56, pls. 21.

No. 13, Irrigation Systems in Texas, by W. F. Hutson, pp. 68,

Is. 10.
3 No. 14, Tests of Pumps and Water Lifts, by O. P. Hood, pp.
1, pl. 1.

oe: 15 and 16, Operations at River Stations in 1897, Parts I
and II, by F. H. Newell, pp. 200.

No. 17, Lrrigation near Bakerfield, California, by C. E.
Grunsky, pp. 96, pls. 16.

No. 18, Jrrigation near Fresno, California, by C. E. Grunsky,
pp. 94, pls. 14.

Geologic Folios: No. 38, Butte Special, Montana, Long.
112° 29’ 30” to 112° 36’ 42”, Lat. 45° 39’ 28” to 46° 02’ 54”, by S.
F. Emmons and G. W. Tower, Jr.

No. 39, Truckee, California, Long. 120° to 120° 30’, Lat. 39°
to 39° 30’, by Waldemar Lindgren.

No. 40, Wartburg, Tennessee, Long. 84° 30’ to 85°, Lat. 36°
to 36° 30’, by Arthur Keith.

168 Scientific Intelligence.

No. 41, Sonora, California, Long. 120° to 120° 30’, Lat. 37° 30”
to 38°, by H. W. Turner and F. L. Ransome.

No. 42, Neuces, Texas, Long. 100° to 100° 30’, Lat. 29° 30’ to
30°, by R. T. Hill and T. W. Vaughan.

No. 43, Bidwell Bar, California, Long. 121° to 121° 30’, Lat.
39° 30’ to 40°, by H. W. Turner.

No..44, Tazwell, West Virginia, Long. 81° 30’ to 82°, Lat. 37°
to 37° 30’, by M. R. Campbell.

No. 45, Boise, Idaho, Long. 116° to 116° 30’, Lat. 43° 30’ to
44°, by Waldemar Lindgren.

No. 46, Richmond, Kentucky, Long. 84°.to 84° 30’, Lat. 37°
30’ to 38°, by M. R. Campbell.

Topographic Folios: Physiographic Types by Henry Gan-
nett.. See this Journal for July, 1898, p. 102. J: Seas

3. A Catalogue of the Cretaceous and Tertiary plants of North
America, by F. H. Know ron, pp. 1-247, 1898, U. 8. Geological
Survey, Bulletin 152.—Mr. Knowlton has furnished paleobotan-
ists a valuakle aid to study by publishing this working catalogue
of the known American species of Cretaceous and Tertiary plants,
without waiting to make it absolutely perfect. Systematic names
are checked back as far as the Kew Index for genera found also
living, and to original source of description for species and the
more important references to later descriptions or illustrations.
Also the matters of geographical distribution and geological
range are given so far as commonly reported in the literature,
while the full knowledge regarding these points is left for future
investigation. The Bibliography is a list of the papers consulted
without assuming that it is a complete list.

4, Iowa Geological Survey. Annual Report, 1897, with accom-
panying papers. SamuEL Carvin, State Geologist, vol. viii, pp.
1-427, plates 1-xxxii, figures 1-13, and six maps, 1898.—This
eighth volume contains detailed reports of the areal geology and,
where ascertained, the faunal lists of fossil species for the counties
of Dallas, Delaware, Buchanan, Decatur and Plymouth, with .
excellent colored maps and frequent half-tone views of important
rock sections and landscapes. H. F. Bain contributes a chapter
on Properties and Tests of lowa Building stones. H. 8. W.

5. A Preliminary Report on a Part of the Gold Deposits of
Georgia, by W.S. YEatss, State Geologist, and 8S. W. McCatiiz
and Francis P. Kine, Assistant Geologists. 1896. Geological
Survey of Georgia, W. 8S. Yeates, State Geologist. Bnuilletin No.
4-A.—This handsome and well illustrated volume of 542 pages
gives an account of the various gold deposits of Georgia, espe-
cially with reference to their present development. As is well
known, these lie chiefly on lines running northeast and southwest,
in the northern part of the State. Some fifty years since, the
amount of gold annually coined at the State mint amounted to
nearly or quite $500,000; but since the exhaustion of the placer
mines, from which most of the gold had been derived, the output
has fallen off very largely, and in 1895 the total production of

Geology and Natural History. 169

gold (and silver) is estimated as but $128,000. The geologist,
however, expresses the opinion that with the deep-mining and
economic working of refractory ores now being introduced, the
future outlook is promising, and the production may be expected .
to reach what it was in former years.

6. The Report of the Governor of Arizona to the Secretary of
the Interior, for the fiscal year ending June 30th, 1898.—This
report, in addition to the discussion of various economic and
agricultural subjects of importance especially to the community
involved, contains also a detailed account of the distribution of
the metallic wealth of the Territory, by W. P. Blake, Territorial
Geologist. How varied and extensive the metallic products of
Arizona are, is well understood, and this fact gives much interest
to this account of the various gold fields, the mines of silver and
argentiferous lead, of copper not the least important of the metals,
also coal, marble and other mineral products. It is interesting
to note that the production of copper in Arizona has been in-
creased fourfold since 1883, and for the first six months of 1898,
amounted to nearly 20 per cent of the entire production of the
United States. |

7. Die natirlichen Pflanzenfamilien, Lieferung 182, 183,
Leipzig, 1898.—Professor ENGLER has carried this important work
nearly to completion. The present installments comprise a portion
of one of the indexes; a general index is to follow at the close.
Certain of the groups of Cryptogams are still in arrears, but they
cannot be much longer delayed. The wealth of illustrations, the
general treatment of orders, and the skill with which minor ref-
erences to distribution and use have been employed, make the
whole work a treasury for general and special botanists. Under
the conditions of publication, it was impossible to avoid the un-
fortunate separation of subjects which should have been united,
but these annoyances are now likely to be forgotten by all the
patient subscribers. There cannot be one of these subscribers
who does not feel that he has obtained far more than the worth
of his money in these clearly printed and attractive volumes.

G. L. G.

8. The Fishes of North and Middle America. A descriptive
catalogue of the species of fish-like vetebrates found in the waters
of North America, north of the Isthmus of Panama; by Davip
STaRR JORDAN and Barton Warren Evermann. Part III, pp.
i-xxiv, 2183-3136. Washington, 1898 (Smithsonian Institution:
Bulletin of the U. S. National Museum, No. 47).—The third part
of this exhaustive Memoir on the Fishes of North America, men-
tioned on page 79 of the last number, has now appeared. Pages
2183 to 2873 are devoted to the description of species, and the
remainder of the volume is given to a key to the families of the
true fishes or Teleostei, the glossary of technical terms and the
index.

wi
oes
a

170 Scientific Intelligence.

Ill. MiscenuANEous ScIENTIFIC INTELLIGENCE.

1. Ostwald’s Klassiker der Hxakten Wissenschaften.—The latest
additions to this valuable series of scientific classics are Nos. 97
to 103, whose titles are given below. We have often had occa-
sion to call attention to this series before, and its value must now
be well appreciated by all of those interested in physics and
physical work. It would be difficult to overestimate the impor-

tance to the physical student of having before him such papers

as are here brought together: They are all of the highest charac-
ter, and a large part of them, as originally published, are com-
paratively inaccessible; but they are here republished in con-
venient form, ready for the student in his study, or the worker
in the laboratory. Space forbids any separate notice of the indi-
vidual memoirs here mentioned, and indeed this is unnecessary,
since they are too well known to require it. Ostwald’s Series
gains additional interest from the fact that an undertaking on
somewhat similar lines is being carried forward in this country
(already noticed in this Journal), as Harper’s Scientific Memoirs,
edited by Professor Ames of Baltimore.

Nr. 97. Sir Isaac Newton’s Optik oder Abhandlung tber Spiegelungen, Brech-
ungen, Beugungen und Farben des Lichts. (1704) II und III Buch.
Nr. 98. Ueber das Benzin und die Verbindungen desselben. Von Hilhard
Mitscherlich. (1834.)
Nr. 99. Ueber die bewegende Kraft der Warme und die Gesetze, welche sich
daraus fiir die Warmelehre selbst ableiten lassen. Von R. Clausius. (1850.)
Nr. 100. Abhandlungen tiber Emission und Absorption von G. Kirchhoff.
I. Ueber die Fraunhofer’schen Linien. (1859.)
II. Ueber den Zusammenhang zwischen Emission und Absorption von Licht
und Warme_ (1859.)
III. Ueber das Verhaltniss zwischen dem Emissionsvermogen und dem Ab-
sorptionsvermogen der Koérper fir Warme und Licht. (1860-1862.)
Nr. 101. .Abhandlungen tiber Mechanische Warmetheorie von G. Kirchhoff.
I. Ueber einen Satz der mechanischen Warmetheorie und einige Anwen-
dungen desselben. (1858.)
IJ. Bemerkung tiber die Spannung des Wasserdampfes bei Temperaturen die
dem Hispunkte nahe sind. (1858.)
III. Ueber die Spannung des Dampfes von Mischungen aus Wasser und
Schwefelsaure. (1858.)
Nr. 102. Ueber physikalische Kraftlinien. Von James Clerk Maxwell.
Nr. 103. Joseph Louis Lagrange’s Zusatze zu Eulers Elementen der Algebra.
Unbestimmte Analysis.

2. University of Tennessee Record of Scientific Engineering.
Published by the University of Tennessee Press, Knoxville,
Tennessee.—Number 7 of this publication, for November, 1898,
pp. 343-405, has been recently received. It contains a number
of original scientific papers, and gives evidence of the scientific
activity at the University of Tennessee. Among the papers we
notice one on Nutrition Investigations, by Charles E. Wait;
several papers on Chemical Analytical Methods; another on
Meteorological Observations by means of Kites, by W. M.
Fulton; also, Notes on the Flora of West Tennessee, by 8S. M.
Bain.

few Store! New Minerals!

Our removal from 64 East 12th Street was completed
and our new store opened during January. Everyone is
charmed with the splendid light and broad aisles which

, make the examination of our great stock of minerals a
pleasure instead of a task. Of course much still remains
to be done in the final straightening out of our stock,
but our salesrooms are now completely arranged and the
filling of orders has once more been resumed. Our facili-
ties for expeditiously handling our increasing business
are greatly improved and every effort will be made to
serve our customers more acceptably in the future than

1% in the past. Our WINTER BULLETIN, giving further an-

Be nouncements and describing recent additions, will be mailed to you if you have

not received it, and we believe it will interest you.

Japanese Quartz Twins. Having traced every quartz twin which has
been exported from Japan. we do not hesitate to claim that the specimens which
we secured in December are the finest which ever left Japan and that no better
ones will come into the market. The large twins are not a new find and substan-
tially all of them have come into the possession of the party from whom we
obtained them, and all of the best of them were purchased by us. The largest
twin is about Bd inches in diameter, another is approximately the same size. Only
_ three of the very large twins now remain, $50, $65, and $75, besides four medium
size ones at $3.50, $5.00 and $12.50.

a _ Other Japanese Minerals. Jatildite, the very rare sulpho-bismuthide of

silver, four specimens at $1.25 to $3.00. Several extra-large nodules of crystal-
lized Arsenic. A series of 20 choice, loose crystals of Stibnite, 25c. te $3.00.

English Chalcosiderite. Over 60 choice specimens of this rare and

attractive mineral at 25c. to $7.50.

The Finest Cleavable Orpiment we have ever seen—a few specimens

only at 25c. to $3.00.

_ Crystallized Hessite. Three extra choice and large specimens, $10, $12.50,

' and $20, recently enriched our stock of smaller specimens at $2.50 to $6.00.

An Enthusiastic Reception has been accorded to our Arkansas Wav-

- ellites. No mineral among all the attractions of our opening sale was so popu-

lar. The specimens are incomparably superior to any we have ever had before.
2c. to $1.00.

_ Serussite from Missouri. Choice crystallized and reticulated specimens,

' 25c. to $5.00, some of them associated with green Pyromorphite.

a. Aurichalcite from Montana. Bright blue, with blue and green Smith-

- sonite, very beautiful; 25c. to $3.50.

. Thousands of other choice specimens, worthy of extended notice, cannot even

be mentioned here.

MINERALS FOR BLOWPIPE ANALYSIS.

i
a This department of our business has developed so rapidly of late that we have
a spared no effort to improve our stock, which now embraces over 300 distinct
a? species, and probably as many more varieties. A complete revised list was pub-
_ lished in our last Fall Bulletin, and we invite the special attention of professors to
it. The facilities of this department are immensely improved in our new store
and we expect to be able to fill orders with greatest despatch.

| GEO. L. ENGLISH & CO., Mineralogists,
Removed to 812 and 814 Greenwich St. (S. W. Corner Jane), New York City.

CONTENTS.

Arr. X.—Contribution to the Study of Gare Me "eins
“phism; by J. M. CusMEnrs 0 5 2g

XI.—Origin of Mammals: by HF. OSBORN» 22 4) eee ‘

XII.—Chemical Composition of Tourmaline ; By 8. ie PEN--
FIELD and H, W. Foorm...2. 0 5... 09333 ee

XIII.—Littoral Mollusks from Cape Fairweather, Patagonia cf
by H. A. Pirspry. (With Platel))-.(- oe |

XIV.—Thermodynamic Relations for Steam; by G.

STARKWEATHER..__ ____22-- A

XV.—Deseriptions of imperfectly known and new Actinians,
with critical notes on other species, II1; by A. EH. Vur-

Pe eae sre gree a eee eg ye ee a eee Me Pe yee er

XVI.—Volumetric Method for the Estimation of Boric Acid ;
by L. C. Jonms 5 2vie2 5023s

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Molecular Masses of Gases, D. BERTHELOT: Preparation |
of Metals and Alloys by means of Aluminum, GOLDSCHMIDT, 154.—Hlectrolytic
Preparation of Beryllium, LEBEAU, 155.—Colloidal Silver, LOTTERMOSER and VON ~
Meyer: Silver peroxide and peroxynitrate, MULDER, 156.—N eodymium, Bou-
DOUARD: Lecons de Chimie Physique professées 4 l’Université de Berlin, J. 1:@
Van’r Horr: Outlines of Industrial Chemistry, F. H. Taorp, 157.—Charge of
Hlectricity carried by the ions produced by Routgen Rays, J. bh TuHomson: Use
of the Coherer in Measuring Electric Waves, 0. BEHRENDSEN, 158.—Color —
Blindness and the Réntgen Rays, BH. Dorn: Regenerating Vacuum ‘Tubes, W. z
Rouins, 159. | és

Geology and Natural History—Age of the Earth, Lorp ae 160. seams
Publications of the U. 8S. Geological Survey, 166.—Catalogue of ‘the Cretaceous |
and Tertiary plants of North America, F. H. Knowiron: Iowa Geological -
Survey, Annual Report, 1897, 8S. CaLvIN: Preliminary Report on a Part of the |
‘Gold Deposits of Georgia, Ww. S. Yeatss, 8. W. McCaLuiE and F. P. Kine, 168.
—Report of the Governor of Arizona to the Secretary of the Interior: Die
natiirlichen Pflanzenfamilien, EN@tur: Fishes of North and Middle Amerioh;
D. 8. JonpAN and B. W. EVERMANN, 169. / Ne.

Miscellaneous Scientific Intelligence—Ostwald’s Klassiker der Exakten weet
schaften: University of Tennessee sae of Scientific Engineering, 170. ;

lished by BENJAMIN SILLIMAN in 1818.

THE

AMERICAN

Epirorn: EDWARD 5S. DANA.

ASSOCIATE EDITORS:

P. BOWDITCH anv W. G. FARLOW, or Camsnivex,

Be WILLIAMS, or New Haven,

ze Proressor GEORGE F. BARKER, or Puinapetpis,
| Proressor H. A. ROWLAND, or Baurimorg,
Mr. J. S. DILLER, or Wasuineron.

FOURTH SERIES.
- VOL. VII-[WHOLE NUMBER, CLVII.}

No. 39.—MARCH, 1899.

WITH PLATES II-y.

NEW HAVEN, CONNECTICUT.
Soon

er by money orders, registered letters, or bank checks.

rEssors GHO. L. GOODALE, JOHN TROWBRIDGE, -

pFussors O. C. MARSH, A. E. VERRILL anv H. 8.

bscribers of countries in the Postal Union. Remittances should

February, 1899.

ILLUSTRATED SUPPLEMENT

TO THE ‘‘COMPLETE MINERAL CATALOGUE” AND

great finds. REVISED POUND LIST.

‘MAILED FREE ON APPLICATION.

ANNOUNCEMENT OF NEW ARRIVALS. |

Sixteen pages. Two handsome photo-engravings of the beautiful stellate and
‘‘spear-head” twin Cerussites, and Anglesite coating same. Brief descriptions
are given of important accessions to stock. Among them are the wonderful —
erystallizations purchased by our Australian collector during his five weeks’ trip
to Broken Hill, which is becoming known as one of the greatest of mineral locali-
ties. Gem-like Raspites and Stolzites, Twin Cerussites, brilliant honey-colored |
Anglesites, Strontianocalcites, Iodyrites, Pyromorphites, 'eto. Gorgeous Tasma- —
nian Crocoites, Franklin minerals (three new species), Endlichites ane other .

THE OLD TRAUTWINE COLLECTION is undergoing rapid disinte-
gration before the stream of orders directed against it.. Collectors were not slow

to accept the invitation extended a month ago. Their desiderata lists are being —

cut down in a most satisfying way, through the purchase of Trautwine specimens.
A full list of the more important is given in the Supplement. If you want long
forgotten and obscure rarities as well as choice examples of commoner species _
found in the early part of the century, write at once. They will be gone in a

few weeks.

THE “NEW SPECIES” LIST,

one of the important features of our “ Complete Mineral Catalogue, ” is Beanent.
up to date in this supplement. Thus you secure free, a careful review, compiled
at considerable expense, of mineralogical discoveries between September, 97, and_
February, 99. It gives name, reference to original publication, form and com-—

position of all new minerals. (The ‘‘Complete Mineral Catalogue,” 186 pp., 40 —

illustrations, with synopsis of Dana’s classification, name, composition and form®
of every known species; supplement to September, ’97, alphabetical index, metal- —
lic classification and other useful tables with Price Lists. Paper bound, 25¢. ; :

cloth, 50c. A work of reference invaluable to mineralogists.) February Supple-

ment is free whether catalogue is ordered or not.

ALL PRICE LISTS FREE. Ng

Dr. A. E. FOOTE,

WARREN M. FOOTE, Manager.

Established 1876.

Mineral Specimens for Scientific and Educational Purposes.

131% Arch Street, Philadelphia, et U.ASs AL

THE

AMERICAN JOURNAL OF SCIENCE

[FOURTH SERIES. ]

1

Art. XVII.—Studies in the Cyperacee; by THEO. Hom.

IX. The genus Lipocarpha R.Br. With nine figures in
the text, drawn by author.

THE genus Jipocarpha was established by Robert Brown

_ upon some plants, which were formerly enumerated by Vahl
as species of Hypelyptum. This latter generic name was,

however, a corruption of “ Hypolytrum” a genus of Richard,
- and since Vahl had included certain species under this, which

did not exhibit the generic character of Hypolytrum as under-
stood by Richard, Robert Brown transferred these to his new

4 genus Lipocarpha: L. argentea, L. sphacelata and L. jili-
* Sormis.

As regards the derivation of the name Lipocarpha, this is

7 not, as stated by American authors, derived from ‘ Aéros fat

and «xdpdos chaff, from the thickness of the inner scales of

some species,” but from “déw fall off, drop,” “from the

whole of its squame being deciduous” as already stated by

q Robert Brown in the Botany of Congo; furthermore the inner

scales are never “fat” but constantly membranateous in the

. species of Lipocarpha, hitherto known. Lipocarpha is a

small genus, containing only seven species according to

- Bentham and Hooker, six of which are described by Beeckeler ;

in later years four more species have been recorded from West
Africa by Ridley. While the genus is chiefly tropical, JZ.

q maculata extends as far north as Virginia, and LZ. Sellowiana
as far south as Buenos Ayres. The habit of the species is

5
;
.

f

a3
Be
‘y

<=
+

*
a.

much the same, reminding one of Ayllinga, and they seem to
prefer damp or periodically inundated ground. Considered

Am. Jour. Sc1r.—Fourts Series, Vou. VII, No. 39.—Marcn, 1899,
12

172 T. Holm—Studies in the Cyperacee.

from a morphological view point, our genus possesses a main
inflorescence composed of a few short and almost roundish
spikes, surrounded by long, green involucral leaves. Each
spike bears a number of spirally arranged bracts, inside of -
which are two minute, membranaceous scales, which surround
a flower. The position of these inner scales is somewhat
singular, being mediane, the lowest one being situated on the
posterior face of rhacheola, the upper one on the anterior face
or a little distance above the supporting bract. Both of these
scales are hyaline, nearly equal in length and apparently of the
same shape, but only the anterior supports a flower. The
question is then to decide whether we have “a one-flowered
spikelet”’ or “a single flower” before us, and this is a point
that has always been so much disputed.

It would seem most natural, however, to define the two
scales and the flower as constituting a one-flowered spikelet,
wherein the lower scale would represent the prophyllon of the
rhacheola and the upper one the bract of the flower. The
position of the lower scale corresponds exactly with that of a
true prophyllon in other Cyperacew. Furthermore its anatom-
ical structure shows us two prominent stereome-bundles and a
distinct bicarinate outline, in contrast to the upper scale, the
bract, in which we have observed a mediane mestome-bundle,
corresponding with other bracts.. The ~
accompanying diagram of a spikelet of
L. maculata shows us the supporting
bract (B), in the axil of which is devel-
oped a rhacheola with a dorsal prophyl-
lon (P) and a bract (0), which supports
the naked flower. This explanation
seems to us the most natural, when we

TCDS es Ree ae consider the minor inflorescences of
Fev a el, Naar ie Fale other Cyperacex, where the rhacheola
nation in the text. is often provided with a basal, empty

prophyllon, bicarinate or tubular as in
Carex, Cyperus, Dulichium, Fuirena and others. Although
Robert Brown did not define the lower scale as a prophyllon,
he considered, at least in his Botany of Congo,* the minor
inflorescence of Lipocarpha as a spikelet, not as a single
flower, a suggestion that was, also, followed by Kunth, Torrey,
Beeekeler and apparently by Baillon. Other authors have
held the opinion that there is “a single flower” instead of a
spikelet, for instance: Vahl, Nees von Esenbeck, Liebmann,
Michaux, Bentham, Ridley, Gray, Schumann, Pax and Goebel.
Pax describes the flower as possessing two prophylla: “eine
hermaphrodite mit zwei medianen Vorblattern versehene

* For references consult the bibliography appended to this article.

7 oa ee =
ai en Cin ~~"

T. Holm—Studies in the Cyperacea. 173

Bliite,’” while Goebel, in his interesting paper on some Cyper-
ace from Java, considers the flower as possessing only one
prophyllon, but is uncertain as to the definition of the other
scale: “Jede Bliite besitzt zwei Schuppen, das eine derselben
wird als Vorblatt zu bezeichnen, das andere ist mir unklar
geblieben.”

As regards the flower itself, we have noticed only one stamen

in Z. maculata and two, seldom three stigmata; purely pistil-

late flowers with rudiments of a stamen were, also, observed.
In this respect LZ. maculata agrees with ZL. argentea, as
described by Goebel, with the exception that the anther of the
stamen in this last species was observed to be two-celled instead
of four-celled, as we invariably noticed in LZ. maculata and L.
microcephala. Concerning the systematic position of our
genus with its “ mediane scales” and “ one-flowered spikelets,”
it seems that Robert Brown was justified in separating it from
Richard’s Hypolytrum, the flower of which is generally
described as possessing “two lateral prophylla.” It is difficult
at present to point out the nearest allies to our genus, con-
sidering that it shows some affinities to Hemicarpha, as sug-
gested by Pax in his treatment of this group for Engler’s and
Prantl’s work upon the natural families. Further researches
on the morphological structure of the other genera may pos-
sibly assist us in determining its correct place within the order,
especially when combined with such peculiarities as may be
observed in the anatomical structure. Im this last respect
Lipocarpha is very interesting, and in order to make our study
as complete as possible, we shall briefly discuss some of the
most important features, which we have observed in L. macu-
lata Kth. from subtropical Florida, L. argentea R.Br. from
Nilagiri in Eastern India and Hongkong, L. sphacelata R.Br.
from the Isthmus of Panama and L. mécrocephala Kth. from
Japan.

The sten above ground,

the scape, is cylindrical and more or less furrowed in the
species examined. The cuticle is thin and perfectly smooth.
The epidermis is here developed into several strata (from two
to four) between the subepidermal bundles of stereome. A
considerable variation is to be noticed in the outer epidermis,
as regards the thickening of the outer cell-walls and the lumen
of the individual cells. Thus we find a strongly thickened
epidermis in JZ. argentea, but one with very thin walls in Z.
microcephala. The epidermis-cells above the stereome are
usually smaller than those surrounding, and contain the char-
acteristic silicious cones, which we have often described in our
previous papers on this order. These cone-bearing cells are

LTA - T. Holm—Studies in the Cyperacew.

thin-walled in Z. maculata and form a projection in the shape
) of a minute papilla (fig. 2), while no other
epidermal projections were observed in any
of the species examined. The outer epi-
dermis-cells, which lie between the stereome-
bundles, are very large and round in trans- —
4 verse sections except in ZL. argentea, in
yas ge Cee which the radial walls have hoon stretched
from epidermis of the so as to render the lumen narrow. It is,
stem of L. maculata. also, to be noted that the epidermis in Z.
argentea forms very prominent ribs on each
side of deep furrows in which the stereome is located. Similar,
but more shallow, furrows are also noticeable in LZ. maculata
and L. sphacelata above the stereome. In ZL. méicrocephala,
however, furrows are, also, present, but the adjoining project-
ing-ribs are not here due to the increased size of epidermis, but
to the stereome. These species of Lzpocarpha thus illustrate a
singular variation regarding the location of the stereome:
being situated in furrows as in LZ. maculata, L. sphacelata and
L. argentea, or in the projecting ribs in L. mzerocephala. The
presence of similar furrows in the stem has, also, been men-
tioned by Rikli in Ficienea, Huirena, Heleocharis and others ;
the stereome, according to this author, is located in furrows in
Feleocharis geniculata R.Br. and Fuzrena, but located in the
projecting ribs in /icinza stolonifera Nees and LF. tenuifolia
Kth. The interior layers of epidermis show a very uniform
structure and the walls are very thin. Stomata are present
and form longitudinal rows between the stereome-bundles ;
they are slightly projecting and the air-chamber is deep and
narrow. In this manner they are freely exposed in our species
with the exception of L. microcephala, where they occupy the
deep furrows between the projecting bundles of stereome.

As regards the mechanical tissue, the stereome, this is very
well represented in the stem of our species, forming relatively
large bundles on the leptome-side of the outer mestome-bundles
or on the hadrome-side of the inner ones. In the first case
the stereome-bundles are situated immediately underneath the
epidermis and border inwards on the chlorophyll bearing pali-
sade-sheath, that surrounds the mestome-bundles. It is gen-
erally thick-walled in our species, especially in L. argentea
and L. sphacelata. When considered in transverse sections, the
stereome forms roundish groups on the leptome-side of the
peripherical mestome-bundles, where it is usually thick-walled ;
the stereome, that supports the inner mestome-bundles, is, on
the other hand, thin-walled and forms an arch around the
hadrome of each bundle. ZLzpocarpha thus possesses a well
developed stereome, which accompanies the mestome-bundles,

T. Holm—Studies in the Cyperacee. 175

but without developing as a continuous ring outside or between
these bundles.

In proceeding to examine the assimilating tissue of the
stem, we notice in Lipocarpha a structure which is very ditf-
ferent from that of other genera, discussed in our former
papers: Carex, Dulichium, Fuirena, Dichromena and
Scleria.* This divergence consists in the very regular arrange-
ment of the palisade-cells radially around the peripherical
-mestome-bundles. In the other genera, previously described,
the palisade-cells were not arranged radially in proportion to
the mestome-bundles, but in proportion to the cross-section of
the stem itself. We notice, therefore, in Lipocarpha that each
of the outer mestome-bundles is surrounded by a closed sheath
of palisade-cells, all of which contain an abundance of chloro-
phyll-grains. A similar sheath of palisade-cells is, also,
observed on the leptome-side of the inner mestome-bundles,
adjoining the arch-shaped group of stereome on the hadrome-
side. The inner mestome-bundles are, hence, surrounded
partly by palisade-cells and stereome. In no other part of the
stem is palisade-parenchyma observable, as the spaces between
the outer mestome-bundles are occupied by the several layers
‘of epidermis.

While this singular arrangement of the palisade-cells in the
stem of Lipocarpha seems to offer an excellent character by
itself, it is, nevertheless, invariably followed by another and
still more remarkable feature, represented by the structure of
the mestome-bundles. We have already mentioned that the
stem possesses two distinct, concentric bands of mestome-
bundles, an outer and an inner. The majority of the mestome-
bundles are situated in the peripherical band and represent two
forms: small and in transverse section orbicular, and larger
which are oval in cross-section. The orbicular are the most
abundant, but they do not all possess a leptome and hadrome
developed to the same extent as the larger, the oval bundles,
‘nor are they all supported by subepidermal stereome, as these
invariably are. Those of the inner band are all large and show
avery pronounced oval outline; besides this, they are only sup-
ported by stereome on their hadrome-side. But all these
mestome-bundles, the large and the small, show the presence
of a typical mestome-sheath, the cells of which are but slightly
thickened in those species which we have had an opportunity
to examine. Inside the mestome-sheath is still another sheath,
the cells of which are conspicuously larger than those of the
mestome-sheath, very thin-walled and filled with chlorophyll.
However the chlorophyll in this sheath differs in a very marked
degree from that of the palisade-tissue by not being differen-

*This Journal, vols. iii, iv, v and vi.

176 T. Holm—Studies in the Cyperacee.

tiated as “ grains,” and by its more intense green color. This
is the sheath which Haberlandt has mentioned as characteristic
of Papyrus cicuta, and which Rikli has so excellently described
after he detected it in other genera of the Scirpoidew. While
Duval-Jouve was well aware of the radial arrangement of the
palisade cells around the mestome-bundles in several species of
Cyperus, as his illustrations show, he does not seem to have
noticed the inner sheath, the “ chlorophyll-sheath,” at least he
does not figure or mention it.*

While the chlorophyll-sheath is completely closed in all the
peripherical mestome-bundles, which, as stated above, are also.
surrounded by an uninterrupted sheath of palisade-cells, it
merely occurs around the leptome in the mestome-bundles of
the mmner band. We remember from the above that these
bundles are not surrounded by palisade-cells excepting on their
leptome-side, and it seems as if the development of the inner,
‘the chlorophyll-sheath,” is somewhat depending on the pres-
ence of palisade-cells; at least the chlorophyll-sheath is strictly
confined to that part of the mestome, which is, also, sur-
rounded by palisade-cells. These structural pecularities: the
development of the assimilating tissue into sheaths around the
mestome-bundles and accompanied by an inner chlorophyll-
bearing sheath, does not seem to depend upon any particular
mode of growth, if we merely consider the stem. We have
noticed this same structure in cylindrical as well as in triangu-
lar stems of certain genera of Cyperacew, in erect as well as
in curved, besides in the singular flattened stems of /ivmbris-
tylis autumnalis KR. et S. It seems, therefore, hardly probable
that the presence of these sheaths depends on any particular

* Professor Warming (Halofyt-Studier, p. 259) insists that Duval-Jouve, not
Haberlandt, should have credit for the discovery of these peculiar sheaths, stat--
ing that Duval-Jouve in the year 1875 described and illustrated these as charac-
teristic of a number of Graminew. However, the chlorophyll-sheath, the one
that lies close up to the mestome-sheath and which borders immediately on the
leptome and hadrome, has so far never been observed in any of the Graminee.
What Duval-Jouve describes and figures is a sheath of palisade-cells, a paren-
chyma-sheath, that surrounds the mestome-bundles and finally a mestome-sheath,
jnside of the parenchyma-sheath. But in none of his descriptions or in his illus-
trations has Duval-Jouve hinted at the existence of a chlorophyll-sheath inside
of the mestome-sheath. It is true, that the mestome-bundles in the leaves of
several Graminee, for instance Cynodon, exhibits three sheaths: one of radially
arranged palisade-cells, a second the ordinary parenchyma-sheath, and a third the
mestome-sheath, but none of these correspond with the sheath which Haber-.
landt detected inside of the mestome-sheath in certain Cyperee, and which Rikli
has lately discussed in his very important paper on this subject. Furthermore
Professor Warming states, that although Duval-Jouve did mention and figure the
palisade-sheath as characteristic of several French Cyperus-species and Galilea,
he, nevertheless, seemed to have overlooked the inner one “ the chlorophyll-
sheath” of Rikli.

While the mere question of priority is a very unimportant one in scientific
research, it must be admitted that Haberlandt was the first to discover this par-
ticular sheath in the Cyperacee.

T. Holm—Studies in the Cyperacee. 177

position of the stems in proportion to the light, and by com-
paring leaves of various genera no apparent connection seems
to exist between their presence and the shape or direction of
the leaf-blades. The genus /imbristylis, for instance, on
which the writer has prepared an article for this Journal,
exhibits a most remarkable variation of growth as regards stem
and leaf, yet invariably possessing the same arrangement of the
ees and development of the inner chlorophyll-bearing
sheath.

Whether the radial arrangement of the palisade-cells around
the mestome-bundles is beneficial to the plant, is a question that
cannot be settled at present. It is, also, very uncertain whether
the dark, green chlorophyll of the inner sheath possesses a
higher assimilating power than that surrounding; very diverse
opinions have been expressed regarding these questions, and a
brief summary may be found in the work of Warming, cited
above.

In returning to the stem-structure of JLzpocarpha, the
leptome and hadrome have attained their highest development
in the inner mestome-bundles, but are not, in any of the species
we have examined, separated from each other by thick-walled
mestome-parenchyma, as described by Rikli (LZ. argented).
There is, furthermore, a large, solid pith which is composed of
thin-walled cells, hexagonal in transverse section, in L. sphace-
lata and L. maculata; in L. argentea and L. microcephala, on
the other hand, we noticed the pith to be interrupted by large
and apparently very irregular lacunes. Cells containing tannin
were observed in the pith, close to the mestome-bundles.

Very different is the structure of rhachis, that part of the
stem which bears the spikelets. The bark-parenchyma con-
tains no chlorophyll and forms a broad ring around the cen-
tral-cylinder; the mestome-bundles are mostly small and
surrounded by a typical, colorless parenchyma sheath and a
thin-walled mestome-sheath, while the inner sheath is totally
absent. The arrangement of the mestome-bundles is, also, dif-
ferent, there being a closed band near the center of the axis,
while a few are seen to be scattered towards the periphery,
entering the rhacheole of the spikelets. The rhachis seems
totally destitute of stereome and the solid pith occupies only a
small part of the central-cylinder. Furthermore we observed
that tannin was present in some of the pith-cells as well as in
the bark-parenchyma.

The leaf.

The stem-leaves of the species of Zipocarpha, which have
been examined, are rather weak, with narrowly linear blades,
closed tubular sheaths, but no ligule. The margins of the

178 T.. Holm—Studies in the Cyperacee.

blade are mostly involute, and the blade, which is often very
asymmetric, is not carinate. The outline of the blade, when
considered in transverse section, is broadly crescent.shaped in
LI. maculata with the two halves of unequal size (fig. 3). The
same is, also, the case with LZ. argentea, but the blade is, in this
species, constricted above the midrib and is thinner in propor-
tion to its width than in the foregoing species (fig. 4). The
blades of LZ. microcephala and L. sphacelata are less asym-
metric and equally thin throughout. The epidermis of the
lower surface of the leaf-blade shows the same development as
in the stem with a single layer outside the mestome-bundles
and two or three rows of cells between these. But on the
upper face of the blade in L. argentea and L. maculata we
find a huge mass of epidermal tissue in layers of considerable
width, from five to six, especially towards the margin; in Z.
microcephala and L. sphacelata, on the other hand, there is but

Fig. 3. Leaf of ZL. maculata, transverse section. x 60.

Fig. 4. Part of leaf of Z. argentea, transverse section. x 60.

one single layer on either face of the blade. In considering
L. maculata and L. argentea the outer layer of epidermis con-
sists of much smaller cells than the inner ones, especially
towards the margins, where two or three subepidermal stereome-
bundles are located. While the outer epidermis in LZ. argentea
is developed as bulliform cells above the midrib (fig. 4), where
no inner layers are developed, we notice in Z. maculata, on the
same place, a tendency of the outer epidermis to develop in
the same way, as bulliform eells, but are separated from the
mesophyll by at least two layers of inner epidermis. In the

T. Holm—Studies in the Cyperacew. 179

two other species the epidermis of the upper face consists of a
single layer of cells of almost equal size, but does not seem to
develop as groups of bulliform cells, so commonly observed in
other genera. The epidermis of the lower surface of the leaf
is composed of cells of nearly equal width excepting those
which cover the stereome; these are usually much smaller and
contain cones of silica. While the lower surface is very even
in L. maculata and L. argentea, we noticed in the two other
species distinct furrows between the stereome-bundles like
those described above as characteristic of the stem. The
stomata are located at the bottom of these furrows, while in
L. argentea and L. maculata they are freely exposed, being on
a level with the surrounding epidermis. As regards the meso-
phyll, this is developed as a sheath around each mestome-
bundle, separating these from the stereome of the leptome-side.
Thus there exists a strong analogy in structure when we com-
pare the bark of the stem with the mesophyll of the leaf,
besides this the mestome-bundles in both show an absolutely
identical composition. The mestome- and chlorophyll-sheath
are developed to the same degree we noticed in the stem, aud
the mestome-bundles are, also, here represented by two forms:
orbicular and oval, when considered in transverse sections
(fig. 5).

Stereome is developed out-
side the larger mestome-
bundles on the lower surface
of the blade, but occurs, also,
in one or two isolated groups
on the upper face and close
to the margin. The margin
itself merely consists of the
two layers of epidermis with
no support of stereomatic
tissue.

The bracts,

Fie. 5. Leaf of L. maculata, showing which subtend the spikelets,
— Ape gis its are membranaceous and con-
with Smee Ea eins of ease ra ealy one rage ty? cpap ig
eee.” 5! 400. ocated in the middle and

forming a slightly projecting
rib on the dorsal face. There is an epidermis of large and
thin-walled cells on both faces of the bract, with the exception
of the margins, where only one stratum of epidermis is devel-
oped. <A few of the cells contain tannin, but no chlorophyll.
The mestome-bundle is surrounded by a single sheath, a
mestome-sheath, the cell-walls of which are but very slightly

180 T. Holm—Studies in the Cyperacee.

thickened. Stereome is present and forms isolated bundles on
the lower surface, one near the midrib, and two on each side, a
short distance from the margins.

The prophylion
of the spikelet shows a very simple structure, there being no
mestome but merely about five ribs of stereome, two of which
are larger than the others so as to render the leaf bicarinate.
Epidermis is developed on both faces excepting near the mar-
gins, where it becomes reduced to one single layer. The epi-
dermal cells are very thin-walled, and none are developed as
bulliform cells. : ,
The bract

of the flower shows nearly the same structure as that of the
spikelet, but much thinner (fig. 6). The epidermis is reduced

Fieé 6. Transverse section of a floral bract of ZL. maculata, showing the mid-
rib and half of the blade; cells containing tannin are drawn with black. x 320.

to a single layer in the lateral parts, from the midrib to the
margins, while the margins themselves possess a double epi-
dermis. Stereome is present, but only on the dorsal face of
the scale covered by epidermis, and tannin-reservoirs were
observed in some of the epidermal cells, indicated with black
in the accompanying figure (fig. 6).
The pericarp

consists of three distinct layers: an outer epidermis of very
thin-walled cells, each containing a silicious cone with minute
warty surface, a few strata of scle-
reids, and finally an inner epi-
dermis. The cells of the inner
epidermis are very thin-walled and
show a rectangular outline in trans-
verse sections (fig. 7). A skeleton
of silica is readily obtained by the
usual preparation, and shows that

Fig. 7. Transverse section of NOt only the cones are silicitied,
the pericarp of L. maculata. Ep = but also a part of the epidermis
the outer epidermis, Scl. = the- ahove these (figs. 8 and O):
Serene tan ince = ™° The roots are very slender and
show a rather open _ structure.

irae
..-!

T.. Holm—Studies in the Cyperacee. 181

There is a thin-walled hypoderm underneath the epidermis, and

the bark is traversed by lacunes of considerable width. The
endodermis has the cell-walls thickened all around, represent-
ing an O-endodermis, inside of which is a pericambium inter-
rupted by proto-hadrome. The innermost part of the root
is occupied by a thick-walled conjunctive tissue with a single
or sometimes two large vessels in the center.

Fig. 8. Epidermis of pericarp of Fic. 9. Same as figure 8, but
L. maculata, seen from above, showing seen from the side. x 560.
the silica-skeleton. x 560.

In considering the anatomical structure of these four species
of Lipocarpha, it seems that our genus possesses several char-
acters by which it may be distinguished from a number of the
other genera, studied so far. The species themselves, may,
also, be recognized in this way, their anatomical characters
being as follows:

D 2 = p =] =|
2 ee qa |e = s S 2)
a | = si ; 5 my 2 al
_ | ~ -_ ra
Bes {et olen Wee fds Bile
& PEE 4 © r= zy hey
EIR a 2 2 = a.|-
o |; o o | a o "3 | o
oa | 4 wD ee o) $622 = on
° is Ors oe) he = gue na > a
Ko) =| o 2 aa hie > om
D ho DS — ie 3 me o =
mv fimeet |) Sia 1S a= mm" = A -
3 ao Ets) - aH = ae
jp The leat. a) es re 2 Ae ae Ati 3
? a. = = =| x "
= mS =p ~ as o ° es ts
oo os UF Lv £5 is — = 2 =o
5) ° at) cS oS =~ = ao eo
=I 2 = ad ° = ae bt os
Siero) saet) | se bs 2. a> 5 = oH On
2S | 2% ie os SH = = cH ie
Be | BS: | So Wee abot a g Oe? | ees
Oe ox eRe) ne | mre = Ai Be as
> D = me] = 3 5
meemacwiata ....--|. — |} — | + }+t|—|—| +/+} —
LL. argentea...---- —- + — +);)—-);4+)/)-);+4+)-
ma sphaceiata ....-|—|—|—+|)/+/—|—}+},—-|+
Be fe Oe ne ee
|
°
ZL. microcephala...) + | — | —! — iy eet ely Ghee
alee ieee atlas RE 2 St So eA

182 T. Holm—Studies in the Cyperacee.

If we finally compare these anatomical characters with those
which Rikli observed in L. argentea, we will notice a few diverg-
-encies. ‘This author states, for instance, that the leaf possesses
one mediane and numerous lateral groups of bulliform cells,
while the leaves we examined showed only one, above the mid-
rib (fig. 4); besides that the lateral parts of the blade were
covered by several layers of large-celled epidermis. Further-
more is stated that the leptome and hadrome are separated
from each other by a zone of thick-walled cells (mestome-par-
enchyma), which we did not notice in any of the species exam-
ined. It seems as if the leaf, which Rikli described, may not
have belonged to L. argenteca, and perhaps not to Lipocarpha
at all. It is, also, to be pointed out that not all of the Cype-
racee, in which a chlorophyll-sheath is developed, have a
scarcity of stereome, since this tissue is well represented in the
species of Lipocarpha. : |

Brookland, D. C., October, 1898.

Bibliography.

Baillon, H.: Monographie des Cypéracées. Paris, 1893, p. 362.
Beckeler, O.: Die Cyperaceen d. Kgl. Herbariums zu Berlin.
(Linnea, vol. xxxvii, p. 112). .
Brown, Robert: Prodromus Flore Nove-Hollandiz. London,
1810, p. 33.

Brown, Robert: Botany of Congo (Miscell. bot. works, vol. i,
London, 1866, p. 144.)

Duval-Jouve: Histotaxie des feuilles de Graminées (Ann. d. sc.
6th series, vol. i, p. 294). :

Duval-Jouve: Etude histotaxique des Cyperus de France (Mém.
d. P Acad. d. sc. d. Montpellier, 1874, p. 347).

Goebel, K.: Ueber den Bau der Aehrchen und Bliiten einiger
Javanischen Cyperaceen (Ann. d. jardin. bot. Buitenzorg, vol.
vii, Leide, 1888, p. 131.)

Haberlandt, G.: Physiologische Pflanzenanatomie. Leipzig,

1884, p. 174.

Hemsley, W. B.: Cyperacez in Biol. Centr. Am. Botany, vol. 3.
London, 1882-86, p. 464.

Hooker, I. D.: The Flora of British-India. Cyperacezx by C. B.
Clarke, vol. vi, London, 1894, p. 667.

Kunth, C. §.: Ueber die natiirlichen Pflanzengruppen der Cype-
reen und Hypolytreen (Physikal-Abhandl. Berlin, 1837, p. 8).

Kurz, 8.: On Pandanophyllum and allied genera, especially those
occurring in the Indian Archipelago (Journal Asiatic Soe. of
Bengal, vol. xxxvili, 1869, p. 70).

Liebmann, F.: Mexicos Halvgres (Kgl. D. Vid. Selsk. Skr.
Kjcebenhavn, 1851, p. 225). | |

Michaux, A.: Flora Boreali-Americana (vol. i, Paris, 1803, p. 29).

oD —
yr ory ee eee

LP. Holm—Studies in the Cyperacee. 183

Nees von Esenbeck: Cyperacez in Martius’ Flora Brasiliensis, vol.
ii, Part I, Wien, 1842, p. 63.

Nees von Esenbeck : Cyperaceze (Acta Acad. Cees. Leop. Carol.
Nat. Cur., vol. xix, Suppl., p. 73).

Pax, Ferd. : Cyperaceze in Engler u. Prantl.: Natiirliche Pflan-
zenfamilien. Leipzig, 1887, p. 106.

Richard in Persoon’s Synopsis plantarum, vol. i, Paris, 1805 »p. 70.

Ridley, H. N.: The Cyperaceze of the West- coast of Africa in
the Welwitsch Herbarium, 1883, p. 162.

Rikli, M.: Beitrage zur vergleichenden Anatomie der Cyperaceen
mit besonderer Beriicksichtigung der inneren Parenchyms-
cheide. Inaug. diss. Berlin, 1895, p. 77.

Schumann, K.: Cyperacee in Engler’s Die Pflanzenwelt Ost-
Afrikas und der Nachbargebiete. Pars C., Berlin, 1895, p.
126.

Torrey, John: Monograph of North-American Cyperaceze (Ann.
Lyceum of New York, vol. iii, 1828-36, p. 288).

Vahl, M.: Enumeratio plantarum, vol. ii, Kj@benhavn, 1805, p.
283.

Warming Eug.: Halofyt-Studier (Kgl. D. Vid. Selsk. Skr. Kje-
benhavn, 1897, p. 259, ete.).

184 G. H. Stone—Granitic Brees of Colorado.

Art. XVIII.—TZhe Granitic Breccias of ae! Peak,
Colorado ; by GEORGE H. STONE.

Geuey PEAK is situated about twenty-three miles southwest
of Leadville, and forms the watershed from which streams
flow northwest into the Roaring Fork and the Grand, south-
west into the Gunnison, and east into the Arkansas. ts ele-
vation is marked by Hayden as 13,956 feet. It is the highest
pinnacle of a very irregular voleani¢ mass consisting of several
radiating ridges one to five miles long.

The eruptive rocks of the district are marked on Hayden’s
Atlas of Colorado as rhyolite. They contain a large amount
of uncombined silica, but sanidine is not abundant, and they
‘more nearly resemble the quartz porphyries of Leadville than
the ordinary sanidine-bearing rhyolites of central Colorado.
Dark silicates occur only in small quantities locally. The
voleanic mass is situated in the midst of a region of schists
containing dikes and irregular bodies of granite.

The volcanic ridges have steep lateral slopes. Cliffs abound,
though in many places the rock disintegrates so as to form a
smooth talus of small fragments. There are a few somewhat
symmetrical cones, but the general outline of the ridges is
rugged, and their tops form a sierra. In several places large
areas are colored bright red, and the local name for the central
mass where Grizzly Peak is situated, is Red Mountain. The
color is due more to the oxidation of pyrite than to the decom-
position of iron-bearing silicates.

Wherever the lava is exposed on the surface, as it often is
toward the tops of the ridges and on some projecting bosses
and crags, it is somewhat scoriaceous and decomposed. But
over most of the higher portions of the ridges the lava is
capped by a layer of breccia composed mostly of fragments of
the lava that are only a little polished and rounded at the angles.
In some places are many subangular fragments of granite or
schists lying upon or among the lava fragments, often being
so abundant as to form all of the surface layer of the breccia
near the top of the mountains. I saw a granite bowlder six
feet in diameter embedded in a solid breccia consisting of lava
fragments.

In some cases the voleanic rock extends to near the bottoms
of the ridges, but in general it forms the upper half or two-
thirds of the ridges. Over the lower portions of the volcanic
ridges and thence extending for a short distance down the slopes
over the adjacent granite and schists, we find a superficial
layer of firmly cemented breccia consisting wholly of rather

G. H. Stone—Granitice Breecias of Colorado. 185

small fragments of schists with some granite. In several
places this is shown by mining tunnels to be fifty and some-
times even more than one hundred feet thick. I found one
tunnel that had penetrated the breccia into granite and schists
in place, in the midst of which is a vertical dike of quartz
porphyry bordered on each side by a vertical layer of breccia
several feet in thickness. This was composed almost if not
entirely of crushed granite and schists. In places the granitic
breccia forms a terrace along the side of the mountain. I
noticed one place where this breccia rests on schists that have
been broken, faulted, and brecciated along the lines of fracture,
so that blocks of schistose rocks twenty to sixty feet in diame-
ter are separated by zones or vein-like bodies of schist-breccia
one to four feet thick. The fragments are only a little rounded,
although different kinds of schists are promiscuously mixed,
showing that there was considerable faulting.

In a paper on the “‘Granitic Breccias of the Cripple Creek
District” * I have described somewhat similar breccias to those
found in the Grizzly Peak district, but on a far larger scale
than the latter. However, the latter are very suggestive when
we study the question of origin.

Obviously when ejectamenta were hurled by violent explo-
sions over a large area adjacent to the volcanic vents, to form
substantially horizontally-bedded breccias or conglomerates, as
was the case in the San Juan mountains, Colorado, and in the
Sierra Blanca range, Lincoln County, New Mexico, the frag-
ments of the older rocks that were ejected from the vents
ought to have been covered and mixed with later ejectamenta
composed of volcanic rocks. But in the case of the dikes at
Cripple Creek above cited, also those at Grizzly Peak, we find
the fragments of clder rocks overlying the fragments of vol-
canic rocks, and these in turn overlying rather narrow dikes
that show but little lateral flow of the lava. The interpreta-
tion is plain. These dikes rose quietly and slowly through the
pre-existing rocks. Before reaching the surface they became
sufficiently solid at their upper extremities to be able to push
fragments of the older rocks before them. In some cases the
latter were at once cemented fast to the tops of the rising
dikes or to the fragments of volcanic rock that were formed
by the breaking up of the thin crust of solidified lava, as the
dike slowly but irresistibly rose above the pre-existing surface.
In such cases they were cemented fast to the rising dike before
they could be washed away. In other cases, as the dikes rose
above the original surface, the fragments of granite, schists, or
other rocks penetrated by the dikes, remained for a time unce-
mented. Under the action of gravity, frost, rain-wash, ete.,

* This Journal, vol. v, Jan., 1898.

186 G. H. Stone—Granitic Breccias of Colorado.

most of these loose fragments then slid down from the tops of
the rising dikes and formed ‘a talus over the lower slopes of the
dikes and thence for a short distance over the adjoining country
rocks. Here they were subsequently cemented into firm brec-
clas.

I see no possible way for explosive volcanic eruptions to
preserve large masses of fragments of the older rocks free
from mixture with fragments of lava, and still less to leave
them overlying all other volcanic matter.

Bowlders of the so-called rhyolite of the Grizzly Peak dis-

trict are fuund in the moraines at Twin Lakes. This proves.
that the voleanic eruption occurred before the glacial epoch..
The position and mode of deposition of the granitic breccias
prove that at the time of the eruption the Sawatch range was
already deeply dissected by erosion, and that there has not
been very much erosion since. .
In the paper above cited I have described the gold-bearing
granitic sand-rock that formed out of the granite crushed by a
rising phonolite dike at the Alhambra mine, Cripple Creek
district. The facts there described show that vein waters,
probably waters heated by the dike which still remained in a
molten condition a few feet below the solidified upper erust,
decomposed the grains of crushed granite, honeycombing the
sand and replacing much of the feldspar with silica and finally
cementing the sand into a firm rock. The mineralized surface
crust of sand-rock was in places charged with many times as
much gold as is contained in any vein that has been found
below in the granite or in the phonolite dike.

Any complete theory of vein deposits will have to treat of the
mineralization of the superficial crusts of breccia that formed
above and along the margins of rising dikes. There are numer-
ous instances of this class of superficial ore deposits found
among the granitic breccias of the Grizzly Peak region, but
the subject requires more detailed treatment than is possible
within the present paper.

Colorado Springs.

Gooch and Austin—Magnesium Phosphate, etc. 187

Art. X1X.—The Constitution of the Ammonium Magnesium
Phosphate of Analysis; by F. A. Goocw and MarrHa
AUSTIN.

[Contributions from the Kent Chemical Laboratory of Yale University—LXXIX.]

IN a recent paper from this laboratory* it has been shown
that the presence of ammonium chloride or other ammonium
salt is necessary in the precipitation of manganese as the ammo-
nium manganese phosphate by microcosmie salt in order that
the precipitate may have the ideal constitution represented
by the symbol NH,MnPO,,.

It was also shown that the solvent effect of the ammonium
chloride upon the precipitated ammonium manganese phos-
phate is not marked when an excess of the precipitant is
present in solution.

The relations disclosed in this paper suggest that the chem-
ical constitution of the precipitate rather than mechanical con-
tamination and varying solubility—the explanations generally
accepted, and, indeed, advocated by one of us in a former
papert—may be responsible for observed variations in the
weight of the residue derived by the ignition of the similar
salt of magnesium, the ammonium magnesium phosphate, pre-
cipitated by an excess of a soluble phosphate from the solution
of a magnesium salt, or from the solution of a soluble phos-
phate by an excess of a magnesium salt.

Precipitation by Kacess of the Soluble Phosphate.

The precipitation of the magnesium salt by an excess of the
soluble phosphate was first studied. For this work a solution
of pure magnesium nitrate was prepared by dissolving the
pure magnesium oxide of commerce in a slight excess of pure
hydrochloric acid and boiling with more magnesium oxide.
After filtering off the excess of magnesium oxide and any trace
of iron or members of the higher groups, the solution was pre-
cipitated by ammonium carbonate, the precipitate was washed
by repeated boilings and filtrations until silver nitrate gave no
precipitate in the solution acidified with nitric acid. This pre-
cipitated carbonate was nearly dissolved in nitric acid and the
solution was boiled with an excess of the carbonate (for the
purpose of removing traces of barium, strontium, and calcium)
filtered, and diluted to definite volume. The evaporation of a
definite volume of the solution and strong ignition of the resi-
due would be a most natural method of establishing a standard

* This Journal, vi, 233.
+ Am. Chem. Jour., i, 391.

Am. Jour. Sor.—Fourts Series, Vou. VII, No. 39.— Marcu, 1899.
13

188 Gooch and Austin—Constitution of the

of the solution, were it not for the fact pointed out by
Richards and Rogers* that the oxide of magnesium retains on
jgnition occluded nitrogen and oxygen enough to increase its
weight sensibly. For this reason the nitrate was converted
to the sulphate and weighed as such—either by evaporating to
dryness in a weighed platinum crucible a definite volume of
the solution, igniting as oxide, and changing to the sulphate
by heating with sulphuric acid; or, by evaporating the mag-
nesium nitrate directly with an excess of sulphuric acid of
half strength. In this treatment the excess of acid was
removed by heating the platinum crucible upon a porcelain
ring or triangle so placed within a porcelain crucible that the
bottom and wails of the inner crucible were distant about one
centimeter from the bottom and walls of the outer crucible.
The excess of acid is easily removed in this way, and the outer
erucible may be heated to redness without danger of breaking
up the magnesium sulphate. The results of this work, taking
O=16, Mg = 24:3, N =14:03, S = 32:06, are given in the
accompanying table.

TABLE I.
MgSO, obtained by MgSO, obtained Theoretical amount
converting ignited MgO directly from of MgO in
into the sulphate. — 50°™3 Mo(NOs)o. MgS0,.
grm. erm. orm.
0°5748 be fs Sea a 0°1924
0°5739 Lae 0°1923
= acts 0°5741 0°1922
Sys 0°5750 0°1925

The magnesium oxide obtained by direct ignition of the
nitrate weighed on the average about 0:0010 grm. more than
the oxide theoretically present in the weighed sulphate from
equal portions of the solution.

Before proceeding to study possible chemical effects of
ammonium chloride in determining the constitution of the
ammonium magnesium phosphate, it is obviously necessary to
define the extent to which the ammonium salt may exert a
solvent action in presence of the precipitant. Fresenius esti-
mated that ammonium magnesium phosphate is soluble in
15293 parts of cold water, but the method of investigation
employed did not entirely preclude the possibility of counting
as ammonium magnesium phosphate soluble material included
and held in the original precipitate.t According to Kissel +
the phosphate, which dissolves in a mixture of ammonia and
water in the proportion of 00040 grams to the liter and in the
proportion of 0:0110 grams to the liter in a similar mixture

* Am. Chem. Jour., xvi, 567.
+ Fresenius, 6% Aufl., p. 805. t Zeitschr. fir Analyt. Chemie, viii, 173.

WE oi GN AEe

Ammonium Magnesium Phosphate of Analysis. 189

containing also eighteen grams of ammonium chloride, is
practically insoluble in the latter mixture if an excess of
magnesia mixture be added; and Heintz* showed that the
effect of adding an excess of sodium phosphate in the solution
is similar.

So far as appears, no quantitative experiments have been
recorded in which the behavior of a mixture of ammonium
chloride and magnesium salt and an insoluble phosphate in a
solution only slightly ammoniacal has been tested, though as a
matter of convenience the use of faintly ammoniacal solutions
and faintly ammoniacal washwater is to be preferred to the
mixture of strong ammonia and water [1:3] ordinarily
employed. Asapreliminary step, therefore, in the work to be
described, experiments were made to find how small an amount
of magnesium could be detected in solution by precipitating with
microcosmic salt, either alone or in presence of ammonium
chloride in faintly ammoniacal solutions. The ammonium
chloride used for these tests (as well as in the similar quanti-
tative work following) was purified by boiling with a faint
excess of ammonia, filtering, digesting twelve hours with

microcosmic salt, and filtering again. The results are given
in Table II.

TABLE II.
Weight of
MgO takenas H(NH,)NaPO,°
the nitrate. 4H.0 taken. Volume. NH,Cl taken Opalescent
grm. grm. cm? grm. precipitation.
0°0003 t45 100 aN marked
00003 ae 500 zl
0°0003 - 100 10 -
00003 ia 500 10 4
0°0003 4s 500 30 faint
00001 ef 100 re marked
0°0001 ae 100 10 re
0°0001 ri 500 10 faint
0°0001 Eo 500 60 a8

The results of these tests show that even so little as 0°0001
grm. of magnesium oxide may be detected in five hundred
cubic centimeters of faintly ammoniacal water containing as
much as sixty grams of ammonium chloride.t It is plain that
strongly ammoniacal liquids are entirely unnecessary in the
precipitation of the ammonium magnesium phosphate under

* Zeitschr. fiir Analyt. Chemie, ix, 16.
+ It was found also, incidentally, that the presence of reasonable amounts of
ammonium oxalate (100 °™’ of the saturated solution) does not interfere with the

precipitation of the ammonium magnesium phosphate by microcosmic salt.

190 Gooch and Austin— Constitution of the

the conditions. In nearly all the experiments to be detailed
use was made, therefore, of faintly ammoniacal solutions and
wash-water.

In Table III are given the results obtained in a study of the
effects of varying proportions of ammonium chloride and the
soluble phosphate upon the constitution of the precipitate.
All precipitates were gathered upon asbestos in the filtering
crucible, washed in faintly ammoniacal water, and ignited as.
usual. In every case the precipitation was practically com-
plete ; for, upon allowing the filtrates with the wash-water to
stand for several days after further addition of microcosmic
salt, nothing but insignificant traces of a precipitate —not ex-
ceeding 0-0001 grm.— ever appeared. In the experiments of
section A precipitations were made in the cold by the action
of microcosmic salt in considerable excess upon the solutions
of magnesium nitrate containing varying amounts of ammo-
nium chloride. In experiments (1) to (5) the liquid was made
faintly ammoniacal after the addition of the precipitant and
the precipitate was filtered off immediately after complete
subsidence; in experiments (6) to (10) the precipitate first
thrown down was redissolved in a very little hydrochloric acid
and reprecipitated by dilute ammonia (the operation being
repeated several times) with a view to improving the erystal-
line condition of the precipitate, and this treatment introduced,
of course, a small amount of ammonium chloride, probably less
than a gram. It will be observed that errors of excess appear
in all of these determinations, those being the greatest in the
experiments in which the largest amounts of the ammonium
salt were present. 3

In the experiments of section B the manipulation was so
changed that the supernatant liquid was poured off (through
the filtering crucible which was to be used subsequently to col-
lect the phosphate) after the precipitate had subsided and the
insoluble phosphate was dissolved in hydrochloric acid and
brought down again, after dilution, by the addition of a faint
excess of dilute ammonia. By thus removing the supernatant
liquid after the first precipitation, the excess of the precipitant
and the amounts of ammonium chloride originally present were
reduced to relatively low limits, so that their effects in the
reprecipitation were at a minimum; and by adding varying
amounts of ammonium chloride, or none at all, before the
reprecipitation, it became possible to demonstrate the individual
effect of that reagent apart from that of an excess of the
microcosmic salt. It will be noted that in experiments (11)
and (12), in which no ammonium salt was added after the
decantation from the first precipitate, the results are ideal, and
that the errors of excess advance as the amounts of ammonium —

Ammonium Magnesium Phosphate of Analysis. 191
salt present in the final precipitation increase. The quantity
of the ammonium salt present during the first precipitation
does not influence the error in the final precipitation unless it
is so large that a simple decantation of the supernatant liquid
would naturally leave an appreciable amount of it to act when

the second precipitation takes place.

TABLE III,
Mg.P.0;,
corresponding Error in Error in HNaNH,PO,.
to Mg(NO:)2 Mg.P.0;, termsof termsof NH,Cl 4H.,0

taken. found. Mg.,P.0;. MgO. present. used. Volume.

erm. grm. erm. erm. orm. erm. cm,

m1) 075311 0°5418 +0°0107 +0 0038 — 2°5 150

{ 2) 0°5311 0°5462 +0°0151 +0 0057 2 150

(3) 075311 0°5408 +0°0097 +0 0035 2 150

( 4) 05311 05500 +0 0189 +0°0068 60 ee 250

(5) 0°5311 0°5520 +0°0209 +0°0075 60 oh 250

(6) 05311 0°5345 + 0°0034 + 0°0012 LF i 150

e7) 0°5311 0°5371 + 0°0060 + 0°0022 es ot 150

( 8) 0°5311 0°5384 +0°0073 + 0°0026 * F 150

moo) 05311 0°5386 +0°0075 +0°0027 ia a 150

(10) 05311 05415 +0°0104 +40°:0037 *€ “ 150
7 -_-"
il) 075311 0°5312 +0°0001 +0°0000 *_ i 150,100
(12) 0°5311 05311 +0°0000 + 0°0600 *_ Vi 150,100
(13) 0°5311 0°5346 +0°0035 +0°0013 242 5 150,100
{14) 05311 05348 +0°0037 +0°0014 242 af 150,100
(15) 0°5311 0°5383 +0°0072 +0°0026 5+5 ae 150,100
(16) 0°5311 05368 + 0°0057 +0 0021 5+5 Ey 150,100
(17) 0°5311 0°5376 +0°0065 +0°0023 10+10 ‘i 200,100
(18) 0°531) 0°5395 +0°0084 +0°0030 10+10 # 200,100
(19) 0°5311 0°5396 +0°0085 +9°0031 60+5 Ms 250,100
(20) 0°5311 0°5389 +0°0078 +0°0028 60+5 re 250,100

It is plain that the errors of excess which appear when

either the ammonium chloride or the soluble phosphate is
present in considerable amount, must be due either to mechan-
ical inclusion on the part of the highly crystalline precipitate,
or to variation in the ammonium magnesium phosphate from
the ideal constitution toward a condition represented by a
phosphate richer in ammonia and correspondingly deficient in
magnesium. If any appreciable amount of the ammonium
chloride present were held by the precipitate, it would natu-
rally be represented by magnesium chloride after ignition, but,
in no one of these experiments, even in those dealing with
sixty grams of ammonium chloride, did the residue, after dis-
solving in nitric acid, give with silver nitrate evidence of the
presence of more than a mere unweighable trace of chloride.
A special experiment, moreover, in which an attempt was

* Probably less than 1 grm

192 Gooch and Austin—Constitution of the

made to determine the silver chloride precipitated from the
solution in nitric acid of an unignited precipitate thrown down
by microcosmic salt in presence of sixty grams of ammonium
chloride, confirms this conclusion : the precipitate was unweigh- |
able. If ammonium chloride present in the solution to so

great an amount is not included in the precipitate in signifi-
cant quantity, it would seem to be unnatural that the micro-
cosmic salt should be included mechanically in any very great
amount. But unless the microcosmic salt was mechanically
included, the increase in weight must be due to the chemical
influence of the reagents—that is, to the production of a phos-
phate rich in ammonium and deficient in magnesium. Ber
zelius* recognized the existence of such a phosphate of
magnesinm; but Wach,t in following the work of Berzelius,.
failed to find it. No attention seems to have been given since
to the existence of such a salt. It would be natural to expect
its formation, if ever, when the precipitating phosphate is in
excess and ammonium salts are present in abundance, with free
ammonia. Obviously the natural effects of all these reagents.
would be toward the production of a salt holding more ammonia
and more phosphoric pentoxide for a given amount of mag-
nesium. The results of the table seem to point strongly to
such tendencies, and, by inference, toward the existence of
such a compound. Thus in experiments (11) and (12), in
which the greater part of this excess of microcosmic salt was
removed by decantation before the second precipitation, while
no ammonium chloride was present excepting the small amount
made by the solution and reprecipitation of the first precipi-
tate, the error is practically nothing. In experiments (13) and
(14), (15) and (16), (17) and (18), all similar to (11) and (12)
excepting that ammonium chloride was present, the average
errors— +0°0036 grm. in terms of magnesium phosphate,
00064 grm., +0:0074, respectively—increase as the ammonium
chloride is increased in the final precipitation. In experiments.
(19) and (20), in which the ammonium chloride amounted to
sixty grams in the first precipitation and to five grams in the
second in addition to the amount that would naturally remain
after decanting the strong solution of the former precipita-
tion, the similarity of this error (+0:0082 in the mean) to
that of the experiments in which smaller amounts of the
ammonium chloride were used throughout goes to show that
only the amount of ammonium salt present in the final pre-
cipitation counts. Further, a comparison of corresponding
experiments of A and B shows very plainly that the treatment
which involves the removal of the large part of the micro-

* Berzelius, J: ahresbericht 3 J ahrgang (1824) tibersetzt von C. G. Gmelin s. 92.
+Schweigger, 1830, Band 29, s. 265.

Ammonium Magnesium Phosphate of Analysis. 198

cosmic salt, the solution of the precipitate, and reprecipitation
tends to reduce the higher indications. Thus, for example, the
error in (2) and (8) is +0°0124 grams in terms of magnesium
pyrophosphate, while in (18) and (14), similarly carried out
except the decantation of the excess of the precipitant, solu-
tion and reprecipitation, the error is +0°0036 grm.

The special influence of free ammonia during precipitation,
was investigated in the following experiments. Definite
volumes of the magnesium nitrate solution were drawn from a
burette into a platinum dish, ammonium chloride—10 grm.—
was added, the magnesium was brought down by dilute
ammonia in presence of microcosmic salt, and strong ammonia
equal to one-third the volume of the solution was added. The
solutions, after standing, were filtered off on asbestos under
pressure in a perforated crucible, and the precipitates were
washed with ammonia diluted to the proportion of three parts
of water to one of ammonia, dried after moistening with a
drop of saturated solution of ammonium nitrate, ignited and
weighed. The results are given in experiments (1) and (2) of
Table IV. In these determinations the mean error reaches
+0-0183 grm. in terms of magnesium pyrophosphate; while
in experiments (3) and (4), made similarly excepting that the
supernatant liquid was decanted from the precipitate first
thrown down, the precipitate dissolved in hydrochloric acid,
and after dilution reprecipitated by dilute ammonia imme-
diately supplemented by enough strong ammonia to make one-
fourth the volume of the entire solution, the error amounts in
the mean to +0°0061 in terms of the pyrophosphate.

TABLE IV.
Mg2P.0,
corresponding Error in Error HNaNH.PO,.
to Mg(NO;)2 MgeP.0;, termsof interms NH,Cl 4H.0

taken. found. Mg.P.0,;. of MgO. present. used Volume

grm. grm. erm. erm. erm. erm em?

(1) 0°5311 0°5503 + 0°0192 +0°0069 10, 225, 200

(2) 05311 0:5505 +0°0194 +0:0070 10 2:5 200
ee ——s*
(3) 0°5311 0°5393 +0:0082 + 0°9029 10, — Deb 200,100
(4) 0°5311 0:5351 +0°0040 +0:0017 10, — 2:5 200,100

In experiments (1) and (2) the precipitate was influenced by an
excess of microcosmic salt, ammonium chloride, and free
ammonia in large amount; in experiments (3) and (4), by
decanting in the manner previously described, by dissolving
the precipitate, and reprecipitating, the effects of an excess of
microcosmic salt and ammonium chloride are reduced to a mini-
mum, and, in a comparison of the results with those of experi-
ments (11) and (12) of Table III the tendency of the free

194 Gooch and Austin—Constitution of the

ammonia comes to view. ‘The results discussed seem certainly
to. point to a general tendency on the part of free ammonia,
ammonium chloride and excess of the phosphate to produce a
salt rich in ammonia and deficient in magnesium, which for a
definite amount of magnesia precipitated must leave upon
ignition a residue weighing more than the normal phosphate.

If it be assumed that a salt of the symbol (NH,),Mg(PO,),—
the next natural step to the normal salt, NH,MgPO,—is present
in the precipitate, the residue which such a salt would leave
upon ignition would be the metaphosphate Mg(PO,),. From
the relations of the symbols for magnesium pyrophosphate
and magnesium metaphosphate the weight of the residue
obtained, and the weight of the pyrophosphate theoretically
derivable from the weight of magnesium salt used, it is possi-
ble, of course, to calculate the proportionate amounts of pyro-
phosphate and metaphosphate present in any ignited residue.
Proceeding in this manner, it appears, that, in order to account
for the variations noted, it is necessary to assume the presence
in many cases of very considerable amounts of the metaphos-
phate. ‘Thus, in the case of those results obtained according
to the usually accepted method of precipitating and washing
with strongly ammoniacal liquids, viz. in experiments (1) and
(2) of Table IV, the proportion of metaphosphate needed to
account for the observed error reaches ten per cent.

Precipitation by Hacess of the Magnesium Salt.

The relations which obtain in the reverse process of precipi-
tation—the action of an excess of the magnesium salt upon a
soluble phosphate—were studied in experiments to be described.
A. solution of pure hydrogen disodium phosphate was prepared
by carefully recrystallizing the pure salt of commerce five
times from distilled water in a platinum dish, dissolving the
erystals, and diluting to definite volume. The standard of the
solution was established by evaporating to dryness in a weighed
platinum crucible known volumes of the solution, igniting the
residue and weighing the sodium pyrophosphate. Magnesia
mixture, the precipitant, was prepared by dissolving fifty-five
grams of magnesium chloride in as little water as possible and
filtering, mixing with this solution twenty-eight grams of
ammonium chloride purified by treating it in strong solutior
with bromine water and a slight excess “of ammonia, filtering,
diluting to one liter, and, after standing for some hours, filter-
ing again.

The tests of the following table show that the precipitation
of a soluble phosphate by the magnesia mixture is practically
complete in faintly ammoniacal solutions even when very
dilute and charged with large amounts of ammonium chloride,

Ammonium Magnesium Phosphate of Analysis. 195

provided the magnesia mixture is present in sufficiently large
eXCESS.

TABLE V.

P.0; in

HNa2PO, Magnesia
taken. mixture. Volume. NH.,Cl.

grm. em?*. cm’, grm. Precipitation.

0°0005 10 100 ns Visible at once
0°0005 50 100 a throughout the liquid.
0°0005 10 100 10 SS
0°0005 10 200 60 rs
00001 a0 250 60 ee
0°0001 10 100 oe i,
0°0001 10 100 10 ef
0°0001 50 200 10 es
0°0001 10 250 60 cs
0:0001 50 300 60 Visible after
0:0001 50 500 60 settling out.

This conclusion was further substantiated by an actual test (by
the molybdate method) of the ignited residue, obtained by evap-
orating a filtrate from ammonium magnesium phosphate.
(equivalent to 0°8614 grams of the pyrophosphate) precipitated
by a faintly ammoniacal solution of magnesia mixture in pres-
ence of sixty grams of ammonium chloride, which gave a
precipitate of ammonium phosphomolybdate yielding 0-0002
grams of magnesium pyrophosphate. It is evident, therefore,
that any considerable deficiencies of weight of the magnesium
phosphate obtained by precipitating equal amounts of a soluble
phosphate by magnesia mixture in presence of varying amounts
of ammonium chloride, cannot be attributed to varying solu-
bility of the magnesium phosphate under changing proportions
of the ammonium chloride. |
The results recorded in section A of Table VI were obtained

by treating definite volumes of the pure solution of hydrogen
disodium phosphate with magnesia mixture, in slight excess
above the amount required to bring down the phosphate, and
making the solution distinctly ammoniacal. After thorough
subsidence, the precipitate was filtered off on asbestos under
pressure in a perforated platinum crucible, washed in water
faintly ammoniacal, dried, ignited and weighed. In experi-
ments (1), (5) and (6), only the ammonium chloride present in
the magnesia mixture was used; in the other cases weighed
portions were added. In the experiments of section B, the
precipitate was dissolved in hydrochloric acid after filtering off
the supernatant liquid, brought down again in dilute solution
by ammonia in distinct excess, and thereafter treated as in the
experiments of section A. The experiments of section C
were conducted similarly to (1), (5) and (6) of A excepting

196 Gooch and Austin—Constitution of the

that the magnesium mixture was introduced into the ammonia-
cal solution of the phosphate drop by drop from a burette.
The precipitations in A, B and C were proved to be practically
complete; for by treatment of the filtrates with more mag-
nesia mixture and standing, no more than a trace—0-0001 grm.
at the most—of the phosphate was found. The ignited resi-
dues never contained more than a mere trace of chlorine.

TABLE VI.
Mg.P.0,7 NH,Cl
correspond- Error in mag- MgCl...
ing to in terms Krror in nesia 6H,0 in
HNaePO, Mg.P.0, ~ of terms of mix- NH,Cl magnesia
taken found. MgeP.0, iP. Volume. ture. added. mixture.
grm erm. erm. erm. em?, erm. grm. grm.
A
( 1) 0°8615 0°8613 —00002 —0-00005 150 1°68 oie aes Bo
( 2) 0°8615 0°8615 0:0000 0°00000 200 = 20 ‘
( 3) 0°8615 0°8602 —0°0013 —0-00036 200 . 20 tt
( 4) 08615 0°8561 —0-0054 —0-°00151 300 a 60 .
( 5) 070852 0:°0862 +0°0010 +0:00028 100 0°28 ate 0°65
( 6) 0°0852 0°0866 +0°0014 +0:00039 100 bd ee <
( 7) 0:0852 0°0847 —0:0005 —0-00014 200 cy 20 9
( 8) 0°0852 0°0830 —0:0022 —0-00062 200 = 20 x
( 9) 00852 00811 —0°0041 —0-00115 300 . 60 af
B
(10) 0°8111 08114 +0:0003 +0°00008 150,100 1°68 lee 3°3
(11) 0°8615 0°8613 —0°0002 —0-00006 150,100 : ag 2 Oh uf
(12) 0°8615 0°8578 —0:0037 —0°00103 200,100 a bay a0 f
(13) 0°8615 0°8487 —0:0128 —0-00358 200,100 - =00 CS
(14) 0°0852 0°0857 +0°0003 +0-00008 100,100 0-28 wee 0°55
(15) 00852 0°0656 +0°0004 +0°00011 100,100 ‘ TQ, 22 i
(16) 0:0852 0°0853 +0°0001 +0°00003 150,100 ne 10,10 et
(17) 0°0852 0°0819 —0-0033 —0:00092 200,100 Fe 20,20
C
(18) 0°8111 0°8071 —0:0040 —0-00112 120 1-4 Een 2°15
(19) 0°8111 0°8052 —0°0059 —0°00165 120 1°4 wit 2°15

While the results are

not entirely regular, the tendency of
the ammonium salt to produce errors of deficiency in propor-
tion to its amount is plain if we compare among themselves
the experiments of A upon similar amounts of phosphate, and
then those of B upon similar amounts of phosphate among
themselves; and by a comparison of corresponding results in
A and B it is clearly shown that the presence of an excess of
magnesia mixture tends to counteract more or less completely
errors of deficiency due to the action of the ammonium chlor-
ide. These facts are quite in harmony with the hypothesis
that the ammonium salt tends to produce an ammonium mag-
nesium phosphate richer in ammonia and phosphoric acid and
poorer in magnesia than the normal salt NH,MgPO,; for,

Ammonium Magnesium Phosphate of Analysis. 197

though the production of such a salt in presence of an excess
of the soluble phosphate compels the combination of a definite
amount of magnesium with more than the normal amounts of
phosphoric acid and ammonia (as was the case in the former
series of experiments), when the supply of the soluble phos-
phate is limited the amount of magnesium associated with it
must fall below the normal (as is the case in the present series
of experiments). Moreover, the behavior of the precipitant is
quite in accord with the hypothesis; for, though the influence
of an excess of the soluble phosphate would naturally tend
(as was observed) in the same direction as that of the ammo-
nium salt and free ammonia, viz., to the production of the phos-
phate deficient in magnesium, the tendency of an excess of the
magnesium salt must obviously be to increase the amount of
magnesium in the phosphate, as was observed in the experi-
ments of Table VI. The hypothesis fits the facts, therefore,
on both sides; and, if precipitation is practically complete (as
was shown to be the case throughout) the argument for the
existence of an ammonium magnesium phosphate—poorer than
the normal salt in magnesium—possibly the salt (N H,),Mg(PO,),
—seems to be strong.

The Practical Determination of Magnesium and Phosphoric Acid.

In determining magnesium by the procedure in ordinary use,
the tendency is strong—as is shown in experiments (1) and (2)
of Table 1V—toward high plus errors, and the error is
due to the combined effects of excesses of the precipitant,
the ammonium salt, and free ammonia. The experiments (11)
and (12) of B, Table ILI, show conclusively that such tenden-
cies to error may be counteracted effectively by pouring off the
supernatant liquid (through the filter to be used subsequently
to collect the precipitate) as soon as the precipitate subsides,
dissolving the phosphate in the least amount of hydrochloric
acid, bringing it down again, after dilution, by a faint excess
of ammonia, filtering (best, we think, on asbestos, under pres-
sure), washing with faintly ammoniacal water, and igniting as
usual.

Many years ago* a method of precipitating the ammonium
magnesium phosphate was advocated by Professor Wolcott
Gibbs, which consists, essentially, in boiling the solution of the
magnesium.salt with microcosmic salt and adding ammonia
after cooling, and by which most exact analytical results were
obtained. Our experience confirms completely that of Dr.
Gibbs, and we desire to direct attention again to a procedure
the advantage of which has, unfortunately, not been broadly
known and accepted. Even in the presence of considerable

*This Journal [3], v, 114.

198 Gooch and Austin—Magnesium Phosphate, ete.

amounts of ammonium chloride this process yields a phosphate
of nearly ideal constitution if only the boiling be prolonged
from three to five minutes. The greater part of the ammonium
magnesium phosphate—about 90 per cent—forms in this process
before free ammonia is added, and the ammonium which enters
the phosphate thus formed is derived from the microcosmic
salt, which must become correspondingly acidic. Under these
conditions, the tendency to form an insoluble ammonium mag-
nesium phosphate richer in ammonia and poorer in magnesia
than the normal salt, does not develop. In the process of Dr.
Gibbs, as well in the modified precipitation process in the cold,
the use of the faintly ammoniacal solution and wash-water is
sufficient and advantageous.

In the precipitation of a soluble phosphate by magnesia mix-
ture the tendency of the precipitant and that of the ammonium
salt are antagonistic, so that the effect of the latter salt is some-
what masked, though manifest. This opposition of effects has
been noted by Mahon,* who, though regarding the actual
attainment of an exact balance as uncertain, ventures the
opinion that accurate results should be attainable by the care-
ful relative adjustment of the proportions of the precipitant
and ammonium salt. Mahon claims to get the best results by
a very gradual addition of magnesia mixture to the ammoniacal
solution of the phosphate containing about sixteen per cent of
ammonium chloride, strong ammonia being added subsequently.
From our observations, however, recorded in section © of Table
VI, it appears that the method of introducing the magnesia
mixture gradually into the ammoniacal phosphate (taken in
quantity sufficiently large to give unmistakable indications)
produces a precipitate deficient in magnesium and so leads to
errors of deficiency in the phosphorus indicated. The use of
strong ammonia, moreover, we have shown to be both unnec-
essary and disadvantageous. Our experiments go to show that
good results may be expected when the solution of the phos-
phate. containing a moderate excess of the magnesium salt and
not more than five to ten per cent of ammonium chloride is
precipitated by making it slightly ammoniacal, the precipitate
being washed in slightly ammoniacal wash-water. In general,
however, and especially when more ammonium chloride than
this proportion, or more magnesium salt than twice the amount
theoretically necessary, is present, it is safer to.-decant the
supernatant liquid from the precipitate (through the filter to
be used subsequently to hold the phosphate), to dissolve the
precipitate in a little hydrochloric acid and reprecipitate by
dilute ammonia, washing with faintly ammoniacal wash-water.

* Jour. Amer. Chem. Soc., xx, 445.

Walker— Crystal Symmetry of the Mica Group. 199

_ Arr. XX.—The Crystal Symmetry of the Minerals of the
| Mica Group; by T. L. WALKER, Assistant Superintendent
Geological Survey of India.

From the time that the symmetry of crystals was deter-
mined exclusively by the aid of goniometrical measurements,
there has been a constant degradation of crystals from systems
of higher symmetry to those of lower symmetry. Many sub-
stances which were once regarded as tetragonal are now looked
upon as rhombic; others that were supposed to be rhombie
have been shown to be monoclinic, while monoclinic has fre-
quently given way to triclimic. The movement has always
been downward, lower symmetry replacing higher. This has
been due in a small measure, to the improvement that has been
made in goniometers, but principally to the progress of physi-
eal investigation of minerals. At first it was considered
remarkable that the optical properties of cubic crystals should
be strikingly different from those of tetragonal and hexagonal
erystals, which in turn differed widely optically from rhombic,
monoclinic and triclinicerystals. At the present time, however,
physical crystallography is looked upon as so important that
no crystal can be admitted to any crystal system till it has
been examined as to its optical, thermal, electric and cohesive,
as well as to its geometrical properties. Only when the physi-
eal and morphological properties agree in every respect, is it
safe to assign a crystal to a particular system.

The degradation caused by placing physical and morpholog-
ical examination of crystals on equal footing, found its climax
in the writings of Mallard,* who seems to have concluded that
all crystals are composed of asymmetric particles which may
by twinning simulate higher grades of symmetry—sometimes
we can detect this twining geometrically or physically, but
even when we cannot detect it, we are to suppose it to exist
ona scale so fine that our means of examination are insuffi-
cient.

The study of etching figures as an indicator of crystal sym-
metry, is one of the newest means discovered by the physical
erystallographer. This method is often so much more delicate
than even the optical method that it points to wholesale degra-
dation. Not only is it much more delicate than the optical
method, but it can often be used when no conclusion can be
obtained by any other means. There have been many protests
against admitting the evidence of this latest and most radical
witness, as is exemplified by Brauns, when he says: “ Wiirde

* &. Mallard, Explication des phénoménes optiques anomaux, Paris, 1877,

200 Walker—Crystal Symmetry of the Mica Group.

man in dieser Weise die Form der Aetzfiguren unter allen
Umstinden als ausschlagend ansehen, so wiirde es bald dahin
kommen, dass man nur noch asymmetrische Krystalle kennt.’’*
It may be urged, however, that etching figures are always in
agreement with the optical properties, and, so far as we can
test them by other reliable methods, they are found to be in
agreement.

Few minerals better show this tendency to degradation from
higher to lower symmetry than do the minerals of the mica
group. The difficulty in determining the crystal system of
the micas by goniometrical measurements is seen in the way in
which very eminent minerologists have differed: Marignae and
Kenngott held that some micas were hexagonal, and others:
monoclinic. Senarmont considered them all as rhombic; the
late J. D. Dana at one time looked upon them as partly hexa-
gonal and partly rhombic; Leydolt found the etching figures
on muscovite cleavage plates to be symmetrical about only one
line, and concluded that the micas were rhombic, but hemi-
morphic about the brachydiagonal. In 1875 Tschermak ob-
served that the optical axis is not exactly at right angles to the
basal cleavage and concluded that the micas were monoclinic.
The only further degradation possible for the minerals of this
group was suggested by Wiik,t+ who observed that the etching
figures on basal cleavage plates are often asymmetric and con-
cluded that at least part of the micas were triclinic. These
results of Wiik, though well known, have not led mineralogists
to look on the micas as triclinic: at present all text-books
describe the micas as monoclinic. The object in writing this
paper is to emphasize the conclusions of Wiik by giving addi-
tional evidence in favor of regarding some, at least, of the
micas as triclinic.

We are accustomed to speak of two classes of micas: Ist,
those whose optical plane is parallel to the clinopinacoid, and
2nd, those whose cptical plane is at right angles to the clino-
pinacoid. As the six-rayed percussion figure has one ray
parallel to the clinopinacoid, we may determine the erystallo-
graphic orientation of crystal fragments by selecting that ray
as clinopinacoidal which is either parallel to or at right angles
to the optical plane. While examining a specimen of rubellan
from the Eifel in this way to determine its orientation, it was
observed that the optical plane was neither at right angles nor
parallel to any ray of the percussion figure but inclined to one
of the rays at an angle of about seven degrees. ‘This led to
wider examination of micas and to the following results.

*R. Brauns, Die optische Anomalien der Krystalle, Leipzig, 1891.
+ Ofvers. Finska Vet. Soc. Forh. 1880.

Walker— Crystal Symmetry of the Mica Group. 201

Phlogopites.

The measurements were made on the stage of an ordinary
petrological microscope and may be considered accurate to
degrees but not to fractions of a degree. The majority of the
specimens examined were cleavage sheets, either without crys-
tal edges or with edges which were too rough for exact
measurement. All those with sharp crystal edges showed that
the edge 001, 010 is parallel to one of the rays of the percussion
figures, so that, where good edges were absent the divergence
of the optical plane from the commonly accepted position
was measured against that ray of the percussion figure which
was nearest to being parallel to it. The divergence for any
one specimen varied with different sheets and for different

arts of the same sheet. Cleavage fragments about half a
millimeter thick often showed no divergence whatever, but
when split to, say a fourth of that thickness, very considerable
divergence was noted, some thin cleavage fragments from one
and the same crystal showing very much more than others of
the same thickness. This, though apparently true in most
cases, is particularly plain in the case of small amber-colored
transparent crystals from crystalline limestone from Upper
Burma, which at a thickness of 1™™ generally show no diver-
gence, but thin plates, about a tenth of that thickness, show
a maximum divergence of 11° 30’. The position of the optical
plane on one of these Burmese crystals is shown in fig. 1, p. 204.
The maximum observed divergences on phlogopites are shown
below.

Maximum
observed.
Locality. Divergence. Angle y*. Material used.

Canada.

2 ee: Sa 9° 45,09 Cleavage fragments.

South Burgess, Ont.._...-..---- 13 61° 53” S %
United States.

GR et Se 17 alae 5 ¥

De Kalb, St. Lawrence Co., N. Y. 6 20’ 61 30 , &

Pierrepont, e ¥s 4 20 ant as a

Rossie, ~ A 2. SO 60 50 ff af

Macomb, ° 1 3 30 63 30 " _

Hammond, . 10 30 ASA Crystals.

Pope’s Mill, Jefferson Co., N. Y. - 7 30 63 Cleavage fragments.

Muscalonge Lake, N. Y..------- 8 30 as Crystals.
Finland.

ot 14° See 4

a as NOSE age Se bs} eS a“
India.

Co 11°30’ 60° 207 ’

oO Se ee 14 63 14 Cleavage fragments.

MEELIS hoe Sato la Sok 3 59 20 as ‘
Ceylon

PMI jt 63° 28” °3 ss

15 62 27 5 ¥

* The angle of the percussion figure opposite the clinopinacoidal edge.—Vide
this Journal, ii, 5, 1896.

202 Walker—Crystal Symmetry of the Mica Group.

These observations prove that in phlogopite the optical plane
is neither parallel to nor at right angles to the clinopinacoid
(so called) and compels us to regard this mineral as triclinic,
which agrees with the conclusions of Wiik, founded upon the
asymmetric etching figures on the basal cleavage plates, as.
shown in fig. 2 [after Wiik]. The fact that thicker cleavage
plates often give no divergence accounts for the common view
as to the position of the optical plane being parallel to the
clinopinacoid. The freedom from divergence in the thicker
plates is to be accounted for by supposing the mica to be twin-
ned polysynthetically in very thin lamellae with (001) as com-
position face, which is the commonest recognised twinning law
for micas, the apparent optical plane in a thick erystal plate
representing the average for the different crystal individuals.
which go to form the cleavage plate. The maximum diver-
gence then becomes the divergence for a sheet which consists
of a single individual ; those peculiar etching figures, which
are somewhat hexagonal in outline with two small inscribed:
equilateral triangles with their points so arranged as to consti-
tute a six-pointed star, would also be explained on the suppo-
sition that one of the triangles represents the etching figures
for one very thin individual, and the other triangle the figures
for the individual immediately below.

Biotite.

It is difficult to determine the ecact position of the optical
plane in the biotites, for the optical angle is so small that the
hyperbolas of the interference figure scarcely separate when
rotated between crossed nicols in convergent polarised light.
This difficulty is also increased by the much darker color of
the biotites and by their giving much poorer percussion figures
against which to measure the divergence in the absence of
crystal edges on the plates. A few of the many biotites ex-
amined gave definite results.

Locality. Divergence. Material used.

India. |

Hazaribagh, Bengal a Cleavage fragments of biotite.
Siberia.

Lake Baikal -..--- 5 ct es fs
Sweden.

Wermland__---_-- oe30" Cleavage fragments of lepidomelane.
Prussia.

DEV Ui(=) VW Rea ae a a 7 Crystal of rubellan.

These observed divergences would refer biotite to the triclinic
system which agrees with the etching figures obtained by Wiik
reproduced in fig. 3.

Lithia Micas.

Only about half a dozen specimens of lepidolite and zinn-

waldite were available for examination. The optical plane is

=.

Walker— Crystal Symmetry of the Mica Group. 2038

sometimes parallel to, at other times at right angles to, one of
the rays of the percussion figure, but never divergent as in the
ease of biotite and phlogopite. ‘This is all the more remarkable,
for the etching figures obtained by both Baumhauer and Wiik
are so very plainly triclinic in symmetry. Wiik’s figures are
reproduced in figs. 4 and 5.

Muscovite.

The optical examination shows no evidence for suspecting
that this mica is other than monoclinic, as we at present regard
it—this agrees with the etching figures of Leydolt, Banmhauer
and Wiik. The optical plane is always at right angles to one
of the rays of the percussion figure, as is consistent with
monoclinic symmetry. Wiik’s etching figures are represented
in fig. 6.

From the above statement it seems evident that part of the
micas are triclinic—biotite, phlogopite, rubellan, lepidomelane
lepidolite, and zinnwaldite, while muscovite is either mono-
clinic, or, if it be triclinic, it is very finely polysynthetically
twinned that we cannot find a triclinic individual large enough
to respond to the optical or etching method. According to
Mallard, these two are the same—monoclinic crystals (so called)
are merely very finely intergrown triclinic twins, the symmetry
plane of the monoclinic being due to the increase of symmetry
which always accompanies twinning. Lévy has suggested that
orthoclase or monoclinic potash feldspar is merely a very
finely twinned microcline. A similar relation may hold for
the micas.

Some observers have referred to the etching figures on the
micas as of monoclinic symmetry. The form of the figures
depends on the reagent used, its strength and the duration of
the action. Though the physical symmetry of the crystals
acted upon is always the same, there may be different amounts
of symmetry revealed by different conditions, just as alum
may be grown in forms which are geometrically holohedral, as
is the case if the crystals are formed in neutral or alkaline
saturated solutions, but if grown from hydrochloric acid solu-
tions, pyritohedral faces are nearly always present, affirming
that all alum crystals are physically hemihedral even when
geometrically holohedral. It would be as unsafe to maintain
that the micas are monoclinic, because under some conditions
certain reagents give etching figures which appear to be mono-
clinic in symmetry, as it would be to conclude that all alum
erystals are physically holohedral because under certain condi-
tions crystals of alum may be obtained which are morphologi-
cally holohedral.

Am. Jour. Sc1.—FourtH Series, Vou. VII, No. 39.—Marou, 1899.
14

204 Walker—Crystal Symmetry of the Mica Group.

I have to thank Professors F. W. Clarke, Washington, D.C.,
S. L. Penfield, New Haven, Ct., G. Tschermak, Vienna, F. J.
Wiik, Helsingfors, and A. Lacroix, Paris, for their kindness in
sending me specimens of phlogopites.

Indian Museum, Calcutta, July, 1898.

EXPLANATION OF FIGURES.

FIGURE 1.—Cleavage plate from a crystal from Upper Burma, showing the posi-
tions of the percussion figure and optical plane.

FIGURE 2.— Etching figures on phlogopite from Pargas, obtained by hydrofluoric
acid. [After Wiik.] *

FIGURE 3.—Etching figures on biotite obtained by hydrofluoric acid. [After
Wiik.|*

FIGURE 4.—Etching figures on zinnwaldite, showing twinning, obtained by hydro-
fluoric acid. [After Wiik.] *

Figure 5.—Etching figures on rose-red lepidolite from Mursinsk, obtained by
action of hydrofluoric acid. [After Wiik.] *

Figure 6.—Ktching figures on muscovite from Kimito, obtained by action of
hydrofluoric acid. [After Wiik.]*

*loc. cit,

if

Sekt SS

Se Wea tre

“
af
}
%
a
@

i Dey met be

A. EF. Verrill—New Actinians. 205

Art. XXI.— Descriptions of imperfectly known and new
Actinians, with critical notes on other species, IV: by
A. E. Verritn. Brief Contributions to Zoology from the
Museum of Yale College, No. LXI.

Family Paractipa. (Continued from p. 146.)
Raphactis Caribea, sp. nov. Figures 17, a, B, page 146.

Base amplexicaul, surrounding completely the branch of a
gorgonian, the two lobes united beneath by suture. The axial
line of the stomodzum lies across the branch, in the type.
Column low, its surface covered by slightly raised elevations,
due apparently to wrinkles, their surfaces smooth; a few small
warts are scattered near the collar, which is thickened and has
about 30 small, short, crowded, longitudinal ridges; their lower
or outer end is largest and most raised. Tentacles numerous,
about 72, crowded in four or five rows, moderately stout,
tapered, entirely retractile. Mesenteries numerous, thin, about
36 pairs join the outer part of the disk, where some of the
smallest bear small gonads. About 12 pairs join the upper
part of the stomodzum, but lower down there appear to be
only six perfect pairs, most of which are fertile, if not all.
Their longitudinal muscles are broad and very thin.

Height of column of contracted specimen, 4"; longest
diameter 11™". Color unknown.

Off St. Vincent, W. Indies, 124 fath., Blake Exped., on a
brown gorgonian.

Stomphia Gosse, 1859.

Carlgren, op. cit., p. 80, 1893.

This genus, established for S. Churchie alone,* has been
studied anatomically and histologically by Carlgren. It is very
nearly allied to what I consider typical Paractis.t It has a

* Paractis vinosa McMur., op.cit., 1894, p. 167. appears to be related to Stomphia.
The number of tentacles is said to be 64, but “the mesenteries are 32 in number,
16 being perfect” (p. 164), and ‘“‘ only the imperfect mesenteries are gonophoric.”’
The number of tentacles given would indicate the presence of 32 pairs of mesen-
teries, which Prof. McMurrich informs me is the case, an error having occurred
in his description. The agreement with Stomphia is therefore close, except that
in our larger examples of the latter the perfect mesenteries are also gonophoric.

Cymbactis feculenta McMur., op. cit., 1893, p. 174, seems also related rather
closely to Stomphia, but has thicker mesogloea, fewer tentacles, 12 pairs of per-
fect mesenteries, and a feebler sphincter than the type.

+1 do not consider Paractis excavata Hert. a true Paractis. The latter name
should be restricted to such plain forms as agree externally with P. impatiens
(Dana). The latter is the first species named by Edw. and Haime and should be
the type if it prove, when anatomically studied, to be a paractid. If not, then
some of the other named species of similar form should be taken as the type. P.
nivea, described below, is such a species, but cannot be taken as the type, for it

206 A. F& Verrill—New Actinians.

smooth wall, much thickened at the submarginal fold. There
are 16 to 24 pairs of perfect mesenteries, the number being
variable; their longitudinal muscles are thin and diffuse. The
perfect and part of the imperfect mesenteries are gonophoric,
in the type. Tentacles are numerous and rather stout.

Stomphia carneola (Stimp.) Ver. Figures 24, 24a—-24d.

Actinia carneola Stimpson, Invert. Grand Manan, p. 7, 1852.

Pen ee Davistt (var. 4) Verrill, Revision Polyps E. coast U. States, pp. 19,

2 Stomphia Churchie Gosse, Ann. Nat. Hist., Ser. III, vol. iii, p. 48, 1859-
Actinol. Brit., p. 222, pl. viii, fig. 5, 1860. Andres, op. cit.,-p. 369, 1884.

Carlgren, Kongl. Svenska, Vet.-Akad. Handl., xxv, 2, p. 80, pl.i, viii, ix, x
(anatomy and histol.), 1892.

2 Stomphia coccinea (Miull.) Carlgren, op. cit., appendix, p. 138 (not Sagartia coc-
cinea Gosse, perhaps not Miuller’s sp.)

Column and base very versatile in form; the base may be
broadly expanded or much contracted; and when detached may
become small and puckered or swollen and conical. Column
may be cylindrical, or almost hour-glass-shaped, or bottle-
shaped, these changes taking place very rapidly. The wall is
smooth in expansion, more or less wrinkled when contracted ;
there are neither verruce nor suckers; a distinct thickened
marginal fold or collar may be formed in contraction. Tentacles
rather long, tapered, moderately stout, versatile, often per-
forated, banded; 96 or more in the larger specimens, in two
or three crowded marginal rows, the outer ones distinctly
shorter and smaller. Disk very changeable, convex or concave.
Mouth often raised; siphonoglyphs two; lips with numerous
(about 14) small folds on each side.

Color variable, usually bright and translucent; column gen-
erally pale pink or flesh-color, irregularly splashed and mottled
with rose red, carmine, or scarlet, except just below the mar-
gin, where there is usually a zone without spots; sometimes
the whole surface is plain flesh-color, or pale greenish white.
Tentacles usually pale pink or flesh-color with three bands of
rose-red or carmine, of which the distal occupies the tip; from
the proximal band a narrow line of red usually runs along each
side of the base to the outer part of the disk, where it may
form a short radial line; but these are often interrupted, so as
to form a circle of red spots on the disk; another circle of
small red spots usually occurs near the mouth. There is
usually a flake-white spot at the inner base of each tentacle.

was not originally referred to the genus by Hdw. and Haime. The same is true
of P. Peruviana (Less.), adopted as type by Andres.

Alloactis, new genus. Type A. excavata (Hert.). Paractide with the circular
muscles of the tentacles unequally developed at the base, it being much thick-
ened on the inside. Wall sulcated. No mesenterial stomata.

A, excavata was taken in 1375 fathoms by the Challenger.

A. &. Verrill—New Actinians. 207

The ground-color of the disk is usually pale pink, but it may
be pale cream-color or greenish white. The mouth is nearly
always surrounded by a narrow circle of bright rose-red,
orange-red, or scarlet on the lips; the angles are usually brighter
scarlet or vermilion; inside of mouth pink or pale salmon.

In transverse section the wall and base are very thin. The
mesenteries are also unusually thin, owing to the very feeble
development of the longitudinal muscles; 16 or more pairs are
perfect and nearly equal in the larger specimens, but the num-
ber is variable; sometimes 24 or more pairs join the upper part
of the stomodzum, or even lower down; but only 16 pairs
usually join it at its proximal end; those of the 4th cycle
are small and often more or less irregularly developed, in
the different sextants. Usually, in the larger specimens, a pair
is regularly developed between all the pairs of perfect ones ;
in the stomodeeal region these are about one-third as broad as
the perfect ones and have longitudinal muscles. All of the
perfect and part of the imperfect ones bear gonads, near the
base. Some small ones of the 5th cycle also occur near the
base. The sphincter muscle is mesodermal, diffuse, elongated,
gradually becoming larger distally, and club-shaped or retort-
shaped. The mesoglcea is considerably thickened in the region
of the sphincter. The two siphonoglyphs are continued down-
ward by a short lobe at each angle of the stomodeeum, which
is large, with about 16 plications on each side.

Height of column, in full expansion, 2 inches (50™"); diam-
eter of disk about 1:5 inches (87™™); length of longer ten-
tacles about ‘6 inch (10 to 15™*).

Bay of Fundy; Eastport Harbor; Johnson’s Bay, near
Eastport, Me. Not rare in 8 to 35 fathoms, on stony bottoms.
Taken by me in 1861, ’63, 64, ’68, ’70, and 1872.

Specimens of this species were described by me in 1864
(Revis. Polyps, pp. 19 (var. 4), 20) as a variety or young of
Urticina crassicornis, which it closely resembles in form and
color. At that time, however, I stated that it agreed closely
with Stomphia Churchie Gosse. The anatomical characters
of the latter were, of course, then unknown.

There can be little doubt that S. Churchie is identical with
our species,* but the latter name is later than carneola, in case
they prove identical when directly compared. It has been
thought by some writers that Actenia coccinea Mill. is identi-
eal with S. Churchie, but of that I cannot judge. Carlgren
also takes this view in the appendix to his article, but accord-
ing to Gosse and others that species is a Sagartia.

* The slight differences between the anatomical description given by Carlgren

and my ownare doubtless due partly to the greater size of my specimens and partly
to the modes of preparation, and to individual variations,

208 A. FE. Verrill—New Actinians.

There can be no doubt of the identity of my specimens
with Stimpson’s carneola. It not only agrees with his deserip-
tion, but Dr. Stimpson was with me on several occasions when
we dredged it at Eastport, Me., in considerable numbers, and
he positively identified it as his species, but he, at that time,
agreed with me that it probably was a variety of UW. crassi-
cornis.

UNO

“a

==
——

\

Explanation of figures.

Fig. 22. Raphactis nitida V. A cluster of nine, surrounding stems of
hydroids. Somewhat less than natural size, from alcoholic specimens.

Fig. 24. Stomphia carneola (St.). Side view of a large specimen, one-half
natural size, from life. 24a, Section of same, with 23 perfect pairs of mesen-
teries, x 2; st, upper part of the stomodzum; d. d, directives. 2406, Longitudinal
section of the marginal region of the wall; m, thickened mesogloea of the collar ;
s, section of sphincter muscle. 24c; 24d, two characteristic forms assumed in
life, when unattached, 4 nat. size.

Fig. 25. Hpiactis prolifera V. One of the types, x2; y, young attached to
sides.

Fig. 29. Sagartia leucolena V. Partly contracted, showing invagination below
margin; a, acontia; from life. ;

All the drawings are by A. H. Verrill, except 24, 24c, 24d, which are from
sketches by the author. .

a
Ae
“a

he

A. E. Verrill— New Actinians. 209

Paractis nivea (Less.). Figure 16, page 146.

Actinia nivea Lesson, Voy. Coquille, p. 81, pl. ili, fig. 8, 1832.
Sagartia nivea Verrill. Trans. Conn. Acad., i, p. 485, 1869 (non Duerden).
Aiptasia nivea Andres, op. cit., p. 175, 1884.

The wall in most parts of the column is very thin, but
strong and parchment-like, often with smooth satin-like ap-
eh and when seen with a lens it may appear somewhat
ibrous transversely ; it shows about 48 mesenterial lines. In
contraction it is variously, but rather closely and deeply,
transversely wrinkled, showing that it is a much elongated
species when in full expansion, The sphincter muscle is
rather broad and diffuse, without a very definite outline. The
mesoglea is thickened near the summit so as to form a fold in
contraction; another fold often occurs lower down, when
strongly contracted, the upper part being somewhat invagi-
nated into the lower. In these examples there may appear to
be a second sphincter, due to the longitudinal contraction.

The perfect mesenteries are usually in 12 pairs, but vary
from 12 to 16, and rarely 24 pairs, all of which appear to bear
gonads. A variable number of small mesenteries occur be-
tween the perfect: ones; usually those of the third cycle are
well developed and bear gonads, but those of the fourth cycle are
very rudimentary and do not bear gonads; small rudiments of
the fifth cycle may also occur. The longitudinal muscles of
the perfect mesenteries are not very thick, but cover most of
their breadth.

Tentacles are very numerous and may be entirely infolded
and concealed; in the larger examples there are over 120;
they are closely crowded in four or five rows; even in alcohol
they are rather long and slender with acute tips; the outer
ones are much smaller than the inner. In some alcoholic
specimens, preserved for thirty years, the tentacles, when they
have been protected by retraction, still retain a reddish brown
tint, but others are pure white. The tentacles in a large num-
ber are fully exposed. Probably the color is variable in life.

The larger specimens are 15 to 18™™ in diameter; some are
25 to 30™™ long, with a diameter of 8 to 10™".

Callao, Peru, on the bottom of an old vessel that had been
more than a year in that port. They were mostly living in
the interstices of large clusters of Discina levis.

This species looks like a Sagartza, but no acontia were found
in numerous specimens dissected, hence I refer it to Paractis.

Subgenus Archactis, nov. Type A. perdia.

Body nearly smooth in expansion, broad, usually cylindrical ;
base broad, muscular; margin without distinct fold or thicken-
ing, in expansion; disk broad, often undulated; tentacles

210 A. EF. Verrill—New Actinians.

hexamerous, very numerous (up to 300 or more) in many rows,
long, tapered, inner ones far from margin, strongly entacmeous.
Mouth large, with two siphonoglyphs and many lateral folds.
Stomodeeum with two large basal lobes, making prolongations
of the siphonoglyphs; mesogloea of wall rather thick, very
flexible, strong ; ectoderm soft; sphincter muscle very broad,
diffuse, mesodermal, moderately thick distally. Mesenteries
hexamerous, very unequal, numerous, with broad, thin, diffuse
longitudinal muscles; 12 or more pairs perfect; all the mesen-
teries fertile, except the directives and the rudimentary ones
of last cycle. |

The following species differs so much from the ordinary
species of Paractis, that it seems desirable to make it the type
of a generic group, especially as the true type of Paractis is
not yet determined. It is closely allied to Stomphza, but lacks
the thick collar and sphincter of that genus, and has a thicker
column-wall, lower down. The mesenteries are also more
graduated in size.

Paractis (Archactis) perdix Ver. Figure 26.

Urticina perdix Ver, this Journal, xxiii, p. 223, 1882; Bulletin Mus. Comp.
Zool., xi, p. 49, pl. vii, figs. 1, la, 1883; Annual Rep. U.S. Fish Comm., xi, p.
534, pl. v, figs. 19, 19a, 19, 1885.

This large and beautiful species has been pretty fully
described as to its exterior, in the works cited above. The
tentacles are about 384 in an average specimen. Alcoholic
specimen of medium size (diameter of column 1°5 to 2 inches)
when dissected had five complete hexamerous cycles of paired
mesenteries, and rudimentary ones of the sixth cycle at the
periphery of the disk (formula, 6+6+12+24448+96 = 192
pairs). Of these, 94, or all belonging to the first five cycles,
except the two pairs of directives, are fertile and bear large
gonads; those of the first three cycles are mostly below the
stomodzeum, while those of the 4th and 5th are in the stomo-
deal region and extend nearly to the disk ; a few of the nar-
row rudimentary mesenteries of the 6th cycle also bear small
gonads, close to the disk. Near the lower part of the stomo-
dzeum there are 12 pairs of perfect thin mesenteries, the six
lateral primaries being distinctly wider and stouter than the
rest. Higher up 12 more pairs (3d cycle) may join the sto-
modeeum and differ but little from those of the 2d cycle ; close
to the disk some of those of the 4th cycle may join the stomo-
deum. Lower down those of the 3d and 4th cycles are free
and successively somewhat narrower, though still broad, while
those of the 5th cycle are much narrower, but all are regular
and fertile. The longitudinal muscles are feebly developed on
all the mesenteries ; even on the perfect ones they are scarcely

A. E. Verrill—New Actinians. 211

4 thicker than the oblique muscles of the opposite side. They

extend over the whole breadth of the mesenteries, in a nearly

uniform thin layer; the mesoglcea is thin with only rudimentary

_ muscle-processes.

_ The stomodzeum is long and large, extending down to about
midheight of the body; the two siphonoglyphs are very deep,

smooth inside, and extend downward as two large pocket-like

lobes, far below the stomodzeum, and nearly to the basal disk.
_ They are about half as long and nearly half as wide as the cen-
_ tral part of the stomodzum. The free parts of the directive
mesenteries are therefore short, narrow, and sterile. The
_ walls of the stomodzum are strongly folded and also have,
each, about 36 principal interior plications on each side, nearly
_ all of which are dark brown in color, even after being 16
years in alcohol.
_ The sphincter muscle is mesodermal, very broad, but not
_ very thick, though it forms more than half the thickness of the
wall distally, where it is slightly club-shaped, in section, and
blunt at the end; it extends down about half the height cf the
column, gradually becoming thinner, but locally thickened
_ where the body is most contracted. The wall is flexible, but
_ strong, and of moderate thickness; the mesoglea is rather
_ thick, compact and nearly even; not much thickened at the
_ collar; the ectoderm is rather thick, soft, and so folded by the
fine wrinkles, running in both directions that the whole surface
often appears to be covered by small wart-like, irregular eleva-
tions and papille, but the mesoglcea does not rise into them.
When living it was smooth, or nearly so, in full expansion.
The whole body contracts very much in alcohol, and more so
than most species. The collar is but slightly marked even in
the contracted specimens, and is not apparent in the live ones ;
the fosse is also very slight; the margin is tentaculate; the
outer tentacles, in alcohol, are very short and conical.

This remarkable species was taken on the Gulf Stream slope,
south of Martha’s Vineyard, in 62 to 192 fathoms. It lives
well in aquaria. One large specimen was kept two months.
On one occasion it caught and swallowed a very active golden
mackerel (Caranz) a foot in length, though several inches of
the fish’s tail projected out of the mouth for some hours.

Synanthus mirabilis Ver. Figure 23.

This Journal, xviii, p. 474, 1879; Bulletin Mus. Comp. Zool., xi, p. 48, 1883,
pl. vi, fig. 9; Ann. Rep. U.S. Fish Com., p. 534, 1885.

Two alcoholic specimens are now figured ; these are united
by sutures, side by side, and together surround a branch of
Paragorgia, \ike a ligature, making a deep constriction, as in

212 A. E. Verrill—New Actinians.

the original types. Tentacles retractile, about 96, rather stout,
closely crowded in four or five rows. Wall smooth, and cov-
ered with a soft ectodermic layer, which rubs off very easily,
leaving a smooth tough surface, slightly suleated near the sum-
mit, and with about twelve principal convergent ridges and
some smaller ones alternating. These do not form very defi-
nite structures, but are apparently due entirely to contraction.
Wall in the region of the collar very thick and firm, with
thick mesoglea. Sphincter muscle mesoglceal, much enlarged
in the collar, gradually thinning out below.

Mesenteries numerous; about 48 pairs may join the outer
edge of the disk, of which some of the smallest bear distally
small gonads. Only six perfect pairs join the middle of the
stomodeeum ; these are fertile at base;* those of the second
cycle are nearly as wide and carry large gonads; those of the
third cycle are narrow, partly fertile; those of the fourth are
rudimentary, except at the outer edge of the disk. The longi-
tudinal muscles of the mesenteries are thin, but cover most of
their breadth; they are largest on the lateral primaries. There
are two pairs of directives and two siphonoglyphs well devel-
oped, but one is deeper. The stomodzumis short. Mouth has
six principal folds on each side. No acontia could be found,
nor any cinclides.

Height, in alcohol, about 4™™; diameter of column 10 to
Ve

Off Nova Scotia, Gloucester fisheries, Lot 534.

The apparent total lack of acontia and cinclides compels
me to place this genus in the Paractide. Its thickened collar
and large sphincter are much like those of Stomphia and
Laphactis. From the last it differs mainly in lacking the
thickened, solid, submarginal ridges. Quite likely larger
examples might have 12 perfect pairs of mesenteries, as those
of the second cycle are wide, in our small specimens. In
Paractide the number of perfect pairs nearly always increases
with age. It also has the habit of Gephyrea, but the latter is
a Sagartian.

* Under the name Paractis lineolata (Dana, sp.) McMurrich has described a
species with only six pairs of perfect mesenteries, which are also sterile, as in
many of the Sagartiade. It had about 96 tentacles and four cycles of mesen-
teries, while Dana’s species had but 24 tentacles, though seemingly of larger
size. The differences combined with the fact that Dana’s species was from the
shore of Orange Harbor, while the Albatross specimens were from N. lat.
8° 16’ 30", in 47 fathoms, render it almost incredible that they can be the same.
The latter may possibly be a Sagartian which had lost its acontia, as often happens.
To avoid confusion I propose to name it Antiparactis dubia, considering the
genus as a doubtful paractid, characterized by the six pairs of sterile perfect mesen-
teries. It has a large mesodermal sphincter muscle in a thickened submarginal

fold of mesogloea, as in many Paractide and Sagartiade. Its base is not
amplexicaul.

=>

A. FE. Verrill—New Actinians. 213

Ammophilactis, gen. nov. Type A. rapiformis (Les.).

Paractid actinians adapted for living in sand, and having a
small base, with a round or obtuse limbus in contraction.
Wall thin, smooth, tough, often translucent and showing the
insertions of the mesenteries. A strong mesoglceal sphincter
muscle, situated in a thickened part of the wall, which, in con-
traction, forms a sort of thickened collar, protecting the sub-
marginal, invaginated region; the latter has a very thin wall,
covered with numerous minute, rounded, adhesive papille,
which are not always visible in expansion, but are capable of
attaching sand grains. Tentacles tapered, numerous (up to
144 or more), in several rows ; the inner ones are far from the
margin, larger, but not much longer than the outer ones.
Mouth large, with two siphonoglyphs, and with 12 to 16 folds
on_each side.

Mesenteries numerous; 24 pairs are perfect, and have broad,
strong muscles. Imperfect mesenteries of the fourth cycle are
also muscular; those of the fifth cycle are feebly developed.
The perfect and the larger imperfect mesenteries bear gonads.

Differs from true Paractis in its elongated body; the re-
duced and feeble base; submarginal band of suckers, ete.

Ammophilactis rapiformis (Les.) Ver. Figures 28, 33.

Actinia rapiformis Les., Proc. Phil. Acad., I, p. 171, 1817.

Paractis rapiformis Edw. & Haime, I, p. 249, L857.

Actinia (?) rapiformis Ver., Revision Polyps, p. 35, 1864; Proc. Boston Soc.
Nat. Hist., x, p. 338.

Paractis rapiformis Ver., Invert. Vineyard Sd., p. 444. [738], 1873; this Jour-
nal, III, p. 436, 1872 (descr.). Dana, Coral Islands, Hd. 2, p. 23, cut, 1874,
Andres, op cit., p. 262.

Sagartia modesta Dana, Coral Islands, Ed. 1, p. 23, figure, 1872 (non Verrill).

A living specimen found by me in a tide-pool at Outer
Island, near New Haven, had the following characters in life:
Height of column in extension, about 5 inches (80™™) ; diam-
eter of disk 1 inch (25™™); of column °5 to 1 inch (12 to 25™™ ;
of base about -5 inch (10 to 12™™); length of tentacles *5 to 1
inch (12 to 25™™).

The longer tentacles are equal to the diameter of the disk,
or even exceed it; they are slender, tapered acute, arranged in
several rather crowded rows, the inner series of six are about
midway between the mouth and margin; the next six only a
little farther outward; the outer ones are only a little shorter,
crowded ; mouth oblong or elliptical with about 14 folds on
each side and a siphonoglyph at each end. Color of column
translucent yellowish white with the mesenterial sutures show-
ing through, as paler lines. Tentacles pale gray, with a
brownish central line and white tips; each has also, on the ©

914 A. EF. Verrill—New Actinians.

inner side, near the end, a flake-white linear median spot or
line, with another like it near the middle, and a less evident
roundish white spot on the inside, near the base; at the base,
on each side, there is a lunate spot of dark brown, the two
nearly meeting in front; from each pair of these spots two
narrow lines of white, edged with orange, run toward the
mouth, but most of them do not quite reach it, but those of
the inner rows reach the lips. The disk is grayish white with
a halo of pale bluish around the mouth.

This specimen was loosely attached to algz and probably
had been recently washed out of the sand by a storm.

Another specimen from near New Haven (coll. J. D. Dana),
of about the same size, preserved in alcohol, has 24 pairs of
perfect mesenteries, nearly equally developed, two pairs being
directives. Between each pair of perfect ones there is a pair
of imperfect ones, about one-third as broad and with strong
muscles. In some places rudiments of another cycle can be
found. The longitudinal muscles of the perfect mesenteries
are strong and extend over most of their breadth, being most
thickened at about the inner third. Gonads are borne both
by the perfect and imperfect mesenteries. The wall is smooth,
thin, but tough, and somewhat parchment-like in alcohol, not
much wrinkled, though sometimes fluted longitudinally; the
mesenteries show through by translucency, as whitish lines.
In contraction, there is a thickened collar below the margin,
containing a rather strong, diffuse, mesoglceal sphincter muscle.
Between the collar and the margin the wall becomes very thin
and soft and is covered with numerous very small adhesive
papillee, or suckers.

The mouth has two siphonoglyphs and about 12 principal.
folds on each side with several other smaller ones, making
about 16. The tentacles number about 120; they form four
or five rows, the outer ones crowded. The marginal tentacles
are smaller than the inner ones, but not much shorter. The
basal disk is small, thin, and without a definite margin, the
limbus being obtusely rounded.

This is one of the specimens figured by Professor J. D. Dana
(Coral Isl., p. 23), but his figure of the mesenteries is entirely
diagramatic and was not made from a section.

Phelliopsis, gen. nov. Type P. Panamensis V. (See p. 144.)

General appearance and habit as in Pheliia. Column much
elongated, but capable of contraction to a short cylindrical,
pyriform, or ovate form and of infolding the summit so as to
conceal the tentacles, though these are often exposed in pre-
served specimens. Base with a well developed adhesive disk,

A. E. Verrill—New Actinians. 215

about as wide as the column. Cuticle of the scapus thin and
closely adherent; integument of alcoholic specimens strongly
and closely wrinkled transversely and longitudinally, but with-
out distinct suckers or verruce. Submarginal zone without
cuticle, plicated in contraction. Tentacles numerous, strongly
entacmeous, arranged in several circles.

Mesenteries numerous, hexamerous, very unequal. Six
pairs of perfect and fertile mesenteries* in the type. These
have very thick and strong longitudinal muscles. Sphincter
muscle rather wide, diffuse. No acontia could be found in
several specimens dissected.

The apparent lack of acontia and cinclides compels me to
refer this genus to the Paractide, though it has the appear-
ance of a Phellia.

Phelliopsis Panamensis Ver. Figures 37, 37a.

Phellia Panamensis Verrill, Trans. Conn. Acad., I, p. 490, 1869. Andres, op.
cit., p. 127, 1884. Hertwig, Voy. Challenger, vi, p. 81, 1883.

The original description of this species is pretty complete,
as to external characters.

The tentacles are in four or five close circles; they vary
from 72 to 96 or more; they are strongly suleated by contrac-
tion. In longitudinal sections the stomodeeum is short, strongly
plicated, and has two siphonoglyphs, while the elongated region
below it is largely occupied by the 12 thick and very muscular,
perfect mesenteries and the large gonads that they bear, extend-
ing from the stomodzum to very near the base (fig. 30, 0).
The imperfect mesenteries, between these, are very much nar-

* In my original description of the type species (1869), I stated that the large
primary mesenteries were very muscular and fertile, but Hertwig (Voy. Chal-
lenger, vi, p. 81, 1883), when little else had been published in regard to the
anatomy of any Phellia, expressed more than a doubt as to the correctness of my
statement. He wrote, in regard to it: ‘‘This so flatly contradicts all observations
on the distribution of the reproductive elements in the Actiniz that Verrill must
somehow have been mistaken.” ‘His observations are of no use for another
reason, namely, that he says nothing about the amon of the septa to the
cesophagus.”

As to the last remark, I originally stated that the fertile septa were the large
ones, “ corresponding to the 12 large inner tentacles” and presumed that every
naturalist would know that such mesenteries always join the cesophagus, or are
“perfect ” to use Hertwig’s term. His remark, however, illustrates the useless-
ness of generalizing as to the internal structure of genera of Actinic that one has
not dissected. This has also been well shown in the case of the species referred
to Sagartia.

As for those actinians that belong to the true Sagartian Phelline, it is now
known, from the studies of Danielssen and others, that in many of the species the
perfect mesenteries are fertile, while in others they are sterile. But it is not
known whether these differences.may not be due to age or to the season of the
year, in many actinians. Perhaps the gonads develop successively on the dif-
ferent cycles of mesenteries in such cases.

The species that Hertwig examined anatomically and described, in the work
cited, as Phellia pectinatus is nota Phellia, but probably belongs to Chondractinia.

216 A. FE. Verrill—New Actinians.

rower and thinner; those of the second cycle bear smaller
gonads, on the upper parts in the stomodeal region. In the
larger specimens seven pairs of small imperfect mesenteries
intervene between the perfect pairs, there being four hexam-
erous cycles. The small mesenteries decrease in breadth suc-
cessively ; those of the second and third cycles have well
defined thickened, but narrow, muscles; those of the fourth
are very narrow.

The wall is flexible, but moderately thick and tough, owing
to the well developed but irregular mesogloea, which rises
internally into irregular lobulate processes, corresponding to
the external plications and wrinkles. It becomes thin and soft
in the capitulum. The circular muscle is distinct and con-
tinuous, but thin. The sphincter in section is elongated,
becoming thicker near the upper edge (fig. 30, s). There isa
large mesenterial foramen in the stomodeal region (fig. 30, 7).

Family Bunopacripa Ver. See No. II, p. 42.

The principal distinction of this family, as compared with
Paractide, is the sharply circumscribed endodermal! sphincter
muscle, which is often only attached to the wall by a part of
its: outer surface. Its outline, in section, is usually ovate or
subcircular, but it varies in form in the same species, accord-
ing to the degree of contraction. Many Phyllactide (Aster-
actis) have a similar sphincter, and it is probable that they will
hereafter be united with this family. The mesenteries and
tentacles may be either hexamerous, decamerous, or octamerous,
and are often irregular in number and arrangement.* There
are usually two pairs of directives and two siphonoglyphs, but
may be only one; sometimes there are three, or even four.
These variations may occur in one species. Perfect pairs of
mesenteries are usually 12 or more, and generally all are fertile
and strongly muscular. Verruce or adhesive suckers are
usually present on the column, but not always. No acontia
are present.

* As a rule, the anatomical investigations of actinians have, up to this time,
been based on a single specimen, or else on a very few, usually collected at the
same season of the year, so that we do not know whether the gonads do, or do
not, develop successively on the different mesenteries, even in common species.
Moreover, in several cases where considerable numbers of specimens of one
species have been dissected, variations often of the most remarkable kind have
been found in the arrangement and number of the mesenteries; number of pairs
of directives, and of perfect mesenteries; form of the sphincter, etc. This is
especially the case in several species of Sagartia, Metridium, Bunodactis, Urticina
crassicornis, etc. Of the latter, I have found many specimens hexamerous, both
as to tentacles and mesenteries; many others decamerous; some octamerous;
and a few irregular or unequally developed on opposite sides. Thus it seems

that such internal anatomical characters are often as variable and no more reliable
than external characters in certain groups of Actinaria.

_

uN Nol y ag

,
Up fp

KZ

HII)
WAN YE

LG

oe.
Ss

Kee

Saud me

Fie. 23. Synanthus mirabilis V. x2; a, b, sutures between two individuals;
c, branch of Paragorgia: d, constricted and broken end of same type

Fig. 26. Paractis (Archactis) perdiz V. 4 size, from life, in extreme expansion.

Fig. 27. Actinostola callosa V. 4 size, from life, type.

Fie. 28. Ammophilactis rapiformis V. % size, from life.

Fig. 30. Stephanauge abyssicola Ver. x2: side view, from an alcoholic speci-
men, showing cinclides.

Fie. 31. The same, partially expanded; type of Actinauge nexilis V.

Fig. 32. Urticina crassicornis, + nat. size; hexamerous variety, from a photo-
graph of a living specimen.

The drawings are by A. H. Verrill; figs. 26, 27, 31 are modified from studies
by J. H. Emerton.

218 uae Verrill—_New Actinians.

Anthopleura Japonica Ver., sp.nov. Figure 39.

The body is contracted in alcohol to an ovate form, the disk
and tentacles being concealed and the base reduced to a small
concave area, but it is somewhat mutilated.

Above the middle the wall is covered with fine transverse
wrinkles and has unequal vertical rows of concave adhesive
suckers. Higher up these are replaced by hollow, round or
conical verruce, and then by more prominent papille, the
upper ones becoming lobulated more or less on the outer and
under sides by small papille. Of marginal papille there are
48 rows, larger and smaller, but of these rows only about 36
are prolonged downward by rows of suckers. Some of the
marginal papille seem to be perforated. The mouth has two
siphonoglyphs. Tentacles about 96, rather stout, tapered,
arranged in three or more crowded rows. Probably they were
rather elongated in expansion. The sphincter muscle is endo-
dermal, rather large, ovate in section, sharply circumscribed.

Mesenteries are in four regular hexamerous cyeles, all much
thickened ; those of the 3d and 4th cycles are successively nar-
rower and smaller; 12 pairs are perfect and nearly equal,
including two pairs of directives. All bear strong, thickened,
pleated longitudinal muscles ; those of the perfect pairs extend
over more than two-thirds of their breadth and are thickest
about the middle. All, or nearly all, the mesenteries bear large
gonads. Some of them are mutilated in the type. The color
in life was not noted.

Simoda, Japan, U.S. N. Pacific Expl. Exped..—Dr. Wm.
Stimpson, 1854. |

Bunodactis Manni Ver., sp. nov.

Column more or less cylindrical; base somewhat expanded.
Tentacles numerous, in several crowded series, tapered, acute.
Column closely covered with numerous, small, rounded ver-
ruce, arranged in many vertical series.

Color of column usually dark green; verruce dark red or
brown; disk, around the mouth pink or light red; tentacles
dark red.

Height, in expansion 25 to 50™™ (1 to 2 inches); diameter of
disk 35™™ or more.

Hawaiian Islands, at and below low water mark in crevices
of rocks and attached to stones.

The above description was prepared many years ago, and
I now find no specimens in our collections to correspond with
it. It was named in honor of my friend, Mr. Horace Mann,
a young botanist who visited the Hawaiian Islands to make
botanical and zoological collections and who contributed largely
to a knowledge of the botany of the islands, but died before
his work was finished.

G. Rh. Wreland—American Fossil Cycads. 219

Arr. XXII.—A Study of some American Fossil Cycads.
Part I. Zhe Male Flower of Cycadeoidea; by G. R.
WIELAND. (With Plates II-IV.)

Introduction.

THE Mesozoic formations of the Rocky Mountain region,
especialiy around the Black Hills, have furnished few fossil
forms of higher scientific interest than the silicified trunks of
Cycads there preserved in great perfection. Most of these
specimens retain both their external features and internal
structure in such minute detail as to make possible a more
complete biological study of the entire group than has hitherto
been attempted. |

There is now in the Museum of Yale University a series of
these rare fossils, including several hundred individuals repre-
sented by complete, or nearly complete, trunks. for this
superb collection of fossil Cycads, Science is indebted to the
generosity and untiring zeal of Professor Marsh, who has also
given the writer the privilege of examining the entire series.

The exact geological age of these Cycads is still in dispute,

‘but their general position isin the middle Mesozoic, and is most

probably Jurassic, as suggested by Professor Marsh in the fol-
lowing statement :

“in the Rocky Mountain region, especially around the margin
of the Black Hills, a definite horizon likewise exists, in which
great numbers of Cycad trunks are found in remarkable pres-
ervation. These Cycads resemble most nearly those from
Maryland, found in what I term the Pleuroccelus beds of the
Potomac formation. In the Black Hills, the age of the hori-
zon has not been accurately determined. * * * Until recently
the Cycads of the Black Hills, although of great size and
remarkable preservation, have not been found actually in place.
In the large collection of Cycads belonging to the Yale
Museum, a few have been discovered apparently where they
grew, and systematic investigation will doubtless show that the
various localities where these fossils have been found around
the Black Hills are all in one horizon. The evidence now
available indicates its Jurassic age, and suggests that it is
essentially the same as that of the Cycad beds in Maryland,
which I regard as a near equivalent of the well-known Cycad
horizon in the Purbeck of England.”’*

* The Jurassic Formation on the. Atlantic Coast.—Supplement.—This Journal,
vol. vi, p. 115, August, 1898. See also, the present number, p. 229.

Am. Jour. Sci.--FourtnH Srries, Vout. VII, No. 39.—Marou, 1899.
15

220 G. fe. Wieland—Americun Fossil Cycads.

Prof. L. F. Ward, however, has recorded his opinion that
this horizon is Cretaceous.* While these fossil Cyeads must
eventually prove of much greater value to stratigraphy than
isolated leaves, as pointed out by Prof. Marsh, their present
interest is mainly of a structural character.

The living Cycads constitute one of the most ornate and
characteristic orders of plants, and, occupying, as they do, a
position on the border-land between the higher Cryptogams
and lower Phanerogams, their ancestral relationship is of con-
siderable interest in itself, though but little progress has yet
been made in working out their phylogeny, owing largely to
imperfect knowledge of the fossil forms.

The Male Flower of Cycadeoidea ingens, Ward.

The male fructification of fossil Cycadean trunks has not
hitherto been determined with certainty. An element of doubt
surrounds the identity of the separate inflorescences usually
held to have been derived from extinct Cycadew, due to the
fact that the Mesozoic Cycads were accompanied by a rich,
varied, and closely related coniferous vegetation. In the ab-
sence of microscopic characters, it becomes a matter of much
difficulty to separate with certainty the fructifications of
Cycads from those of some Conifers.

Previous observations on this subject are hence quite limited.
Of the supposed fossil Cycadean male fructifications, the pyri-
form axis described by Williamson’ has given rise to the most
discussion. While it is not necessary to review the literature
on this subject here, it may be mentioned that Seward,” after
declaring his belief that the Wealden examples of Williamsonia
are generically identical with Lennettites, writes as follows :—
“As regards the question of maleand female inflorescence, I am
unable to recognize any sexual differences in the various
examples from the Wealden beds, and there does not seem to
be any good reason for regarding the so-called male Wellzam-
sonias among the Jurassic specimens as in any way proved to
be of that nature. In comparing Weallcamsonia with Bennet-
tites we have to rely solely on the female inflorescence of the
latter plant, and it would seem, so far as our present evidence
goes, we have more reason for speaking of Wallzamsonia as
the female inflorescence. As to the nature of the male
inflorescence we are still without satisfactory evidence.” As
will be shown later, the present investigations add a strong
degree of probability to this view, if they do not indeed
demonstrate its correctness. The reviews of Solms-Laubach,’
and of Schimper’ who alludes to the male axis of Wellcamsonia
as a Pandanus-like form, may also be consulted.

* The Cretaceous Rim of the Black Hills.—Journal of Geology, vol. ii, pp.
250-266, 1894.

G. R. Wieland—American Fossil Cycads. 221

_. The principal loose fossil fruits usually referred to by paleo-
botanists as male fructifications of Cycadean origin are Andro-
strobus Balduini, Saporta,* and Zamites familiaris, Corda.
Still other and more problematical forms are mentioned by
Seward,“ and Solms-Laubach.* In no ease, however, is the
microscopic structure preserved, or a knowledge of these forms
at all complete. There is, therefore, no direct basis for com-
parison between -these conical fruits and the capsular male
fructification of Cycadeoidea here discussed.

~-EX~. -.---- Lg

>
~ = 4
- > ,

Male flower bud of Cycadeoidea ingens, Ward.
FiguRE 1.—Diagram of longitudinal section.

FIGURE 2.— e transverse
a, group of sori; 6, fleshy outer wall; c, central cavity; d, a sorus; e, bract
hairs; f, undetermined body; g, basal:sorus; hf, receptacle; 7, involucral bract ;

k, peduncle.
Both figures are natural size.

The type specimen of Cycadeoidea ingens, Ward,” which
bears the beautiful flower bud here described, is the type of
the species, and is a magnificent trunk, nearly perfect, weigh-
ing 304 kilograms (671 pounds). In Plate II, it is represented
with the flower bud still in position near the summit, where it
appears as a conelike projection 3°5 cm in height. Sections
of this bud are shown in Plates II] and IV. The trunk as
seen in Plate II is inclined slightly forward to display the
bases of a number of other buds of flowers equally advanced
in growth, but not wholly preserved.

229 G. LR. Wieland—American Fossil Cycads.

The main interior structures of this specimen are indicated
in the diagrams figures 1 and 2, page 221, based respectively on
the photographs of sections shown in Plate III, and in Plate IV,
ficure 1.

BE cleo this flower bud and the apex of the trunk, distant
15 em, is a thickly-set mesh of stipulary chaff enveloping a
series of finely-preserved, just emergent young leaves, which
will be described later.

As removed from the trunk for examination, the general
shape of the flower bud was subcylindrical, with a slightly
tapering base and a strongly tapering summit. The length,
including a considerable portion of the peduncle, was 73 mm,
the basal, middle, and apical diameters being 35, 45, and 25
mm, respectively. . Laterally it was covered with a series of
imbricating involucral bracts, the summits of which had been
eroded away. One very noticeable point was the division of
the free summit into twelve subequal sectors by radiating
lines apparently due to differentiation of tissues.

The upper portion of the peduncle, figure 1, #, is seen to be
traversed by a series of fibro-vascular bundles, many of which
pass off into the involucral bracts. The latter are covered by
a mat of fine hairs quite similar to those borne by the involu-
eral bracts of Bennettites.” A number of much larger and
coarser hairs also arise directly from the peduncle, between the
bases of the bracts. The position of the receptacle is indicated
centrally by nearly clear quartz, lacking well-defined structure,
and peripherally by a shoulder-like offset bearing most of the
involucre of bracts.

The basal sori, figure 1, g, are not fully developed, and are
irregularly distributed in pockets. Following this absence of
arrangement, there is a soral grouping into eleven planes, which
successively increase their angle to the floral axis until the
terminal series rises vertically. It should be noted, however,
that this arrangement is only apparent in longitudinal radial
sections, situated at regular intervals with reference to the
plan of the flower. In other sections, the sori are crowded
together without a traceable arrangement, and the grouping in
planes is obscured.

Inside the fleshy spore-bearing portion of the flower, which
may for convenience be termed the soriferal axis, is a central
cavity of quite regular form, lined by a druse of quartz crystals,
figure 1, ¢, in places resting on a thin layer of chalcedony, which
in turn rests on the silicified tissue.

At f, figure 1, is shown an undetermined cylindrical body
which has its origin on the receptacle, and curves upward round
the base of the soriferal axis, passing half-way up its side. As
there are at least ten, if not twelve, of these bodies quite

G. R. Wieland—American Fossil Cycads. 923

regularly distributed in the same relative position, they are in
some way connected with the floral plan, though not likely as
essential organs.

In figure 2, which is a diagrammatic sketch of the transverse
section through the middle of the soriferal axis, based on the
photograph shown in Plate IV, figure 1, the plan of the flower
is clearly shown. There is first a series of imbricating bracts,
the thickly-set hairs of which are not indicated in the sketch.
These are in cross section quite thin and scalelike in the
middle and upper region of the soriferal axis, especially the
interior series. A cross section of the bracts near their origin
is seen in Plate IV, figure 2, which represents a section through
the base of the flower, about 2 mm below the basal sori. At
this height, the cross section of the bracts is approximately of
the same form as that of the petioles of the trunk itself.

Inside the involuere of bracts, figure 2, b, is the fleshy integu-
ment surrounding the cross area of soral distribution, which is
approximately that of a twelve-rayed star, with short points
and a hollow center. This is a second clear indication that
there is a general plan of twelve, the radial lines of the summit
previously mentioned being the first. The sori tend strongly
to arrange themselves in tandem pairs parallel to the sides of
the twelve points, thus forming an approximation to twelve
radial series of from four to six parallel V-shaped groups.
Only the examination of a series of successive thin sections
now being made will show how far the soriferal surfaces are
barren or depart from this arrangement.

It is obvious that the fleshy onter wall extends inward as
a homogeneous tissue to form most, if not all, of the soriferal
surface. Whether or not these surfaces had any connection
with a central axis is obscured by the presence of the central
cavity mentioned above. There does not, however, seem to
have been any such connection. That there is in any case a
series of surfaces which must be regarded as the homologues
of sporophylls is a point on which botanists will doubtless agree.
The discoveries of [keno and Webber,” and the investigations
of Lang,” do not leave any doubt as to the correct terminology
of Cycadean inflorescence.

The finer structure of these organs must be treated at
ancther time, though it should be explained now that the sori
are composed of from twenty to forty clavate sporangia. Each
sporangium is about 1 mm in length and 54; mm in diameter,
and contains 500 or more rounded or subangular bodies not yet
completely studied, but tindoubtedly the spores or pollen grains.

With regard to the homology of these structures, several

facts are worthy of mention. The radial divisions occurring

294 G4. RR. Wieland—American Fossil Cycads.

on the summit are found to persist for a considerable distance
downward, and under the microscope are seen to be due to the
presence of two lignified layers, a single cell in thickness.
They correspond to the twelve vertices of soral distribution
mentioned above, and their presence is against the idea that
the soriferal axis is derived from the fusion of the sporophylls
of a male cone like that, for instance, of Zamza integrifolia.
Another and much more tenable hypothesis is that the soriferal
axis is a series of twelve fused leaves with their sorus bearing
pinnules turned inwards.

While it is evident that the inflorescence under consideration
represents a development far in advance of that of existing
Cycads, the most striking fact observed is the linear distribu-
tion of the sori. This is an archaic character most marked
in the Carboniferous ferns, especially the J/arattiacee, and
their allied forms, living and extinct.

It has probably been the opinion of all botanists, since it
was discovered that Stangeria paradowa was a Cyead and not
a tree fern, as originally described, that the relationship between
the Ferns and Cycads must be an exceedingly close one. All
later investigations have tended to strengthen this belief.
Scott” has recently stated that the evidence in favor of the
filicinian ancestry of the Cycadee must now be considered
overwhelming.

It was scarcely to be expected, however, that forms bearing
strong testimony on this point should display such a marked
combination of advanced as well as ancestral characters.

The further study of the Yale collection promises to afford
many new details regarding Cycadean structure and affinity.
In addition to the facts presented in this paper, it has been
now determined for the first time that the leaf characters of
Cycadeoidea were approximately those of Zamza and Doon,
and had the prefoliation of these genera. Moreover, the
ancient forms, like the existing Cycads, were dicecious.

In closing, the writer wishes to express his best thanks to
Professors Marsh and Beecher for the indispensable aid he has
received during the entire course of the present investigations.
He is also indebted to Professor Lester F. Ward for valuable
_ suggestions and reference to important literature.

Yale Museum, New Haven, Conn.,
February 20, 1899.

G. R. Wieland—American Fossil Cycads. 225

References to literature.

1. Corda, August J.—Pflanzen. Theil H. in Reuss’s Verstein-
erungen der bémischen Kreideformation, pp. 81-87, Tafel xlix,
figures 10-13c, Stuttgart, 1845-46.

2. Williamson, W. C.—Contributions toward the History of
Lamia gigas. Transactions of the Linnean Society of London,
vol. xxvi, pp. 663-674, plates lii and lii, London, 1870.

- $8. Carruthers, William.—On Fossil Cycadean Stems from the
Secondary Rocks of Britain. Trans. Linn. Soc., London, vol.
XXvl, pp. 675-708, plates liv—lxiii, London, 1870.

4. Saporta, G. de.—Palzoutologie francaise. Plantes juras-
‘siques, vol. li, p. 209, plate cxv, figures 1 and 2, 1875.

5. Eaton, Daniel C.—Ferns of North America, 2 vols., Bos-
ton, 1879-80. 7

6. Nicholson, George.—lIllustrated Dictionary of Gardening,
London, 1887-89.

7. Schimper, W. Ph., und Schenk, A.—Paleophytologie (II.
Abtheilung, Zittel’s Palzeontologie), pp. 211-232, figures 158-166,
Miinchen und Leipzig, 1890.

8. Solms-Laubach, H. Grafen zu—Ueber die Fructification von
Bennettites Gibsonianus Carr. Botanische Zeitung, pp. 789-798,
805-815, 821-833, 843-847, Tafeln ix und x, Leipzig, December,
1890.

9. Fossil Botany (rans by Garnsey; revised by
Balfour), pp. 85-100, figure 1, Oxford, 1891.

10. Capellini, Giovanni, e Solms—Laubach, Conte E.—I tronchi
di Bennettites dei Musei Italiani, Notizie storiche geologiche,
botaniche. Memorie della Reale Accademie delle Scienze, Insti-
tuto di Bologna, serie v, tomo ii, pp. 161-215, tavole i-v, Bologna,
June, 1892.

11. Ward, Lester F.—The Cretaceous Rim of the Black Hills.
Journal of Geology, vol. ii, pp. 250-266, April-May, 1894.

12. Lignier, Octave.— Végétaux fossiles de Normandie. Struc-
ture et Affinités du Bennettites Morierie Sap. et Mar. (sp.), 78
pages, vi plates, Caen, 1894.

13. Campbell, Douglass H.—The Structure and Development
of the Mosses and Ferns (Archegoniatz), London, 1895.

14. Seward, A. C.—Catalogue of the Mesozoic Plants in the
Department of Geology, British Museum (Natural History). The
Wealden Flora. Part Il. Gymnosperme, pp. 1-182, figures
1-9, plates i-xv, London, 1895.

£5, On Cycadeoidea gigantea, a new Cycadean Stem
from the Purbeck Beds of Portland. Quarterly Journal of the
Geological Society of London, vol. lili, pp. 22-39, figures 1-4,
plates i-v, London, February, 1897.

16. Kerner von Marilaun, Anton.—Pflanzenleben, Leipzig, 1896.

17. Webber, Herbert J.—Peculiar Structures occurring in the
Pollen Tube of Zamia. Botanical Gazette, pp. 453-459, plate xl,
June, 1897.

226 G. R. Wieland—American Fossil Oycads.

18. The Development of the Antherozoids of Zamia.
Bot. Gaz., pp. 16-22, figures 1-5, July, 1897.

19. Notes on the Fecundation of Zamia and the Pollen
Tube Apparatus of Gingko. Bot. Gaz., pp. 225-235, plate x,
October, 1897.

20. Scott, D. H.—The Anatomical Characters presented by the
Peduncle of Cycadacee. Annals of Botany, vol. x, pp. 399-419,
plates xx and xxi, London, August, 1897.

21. Lang, William H.—Studies on the Development and Mor-
phology of Cycadean Sporangia. Ann. Bot., vol. xi, pp. 421-438,
plate xxii, London, September, 1897.
see 2: Worsdell, W. C.—The Vascular Structure of the Sporo-
phylls of Cycadaceze. Ann. Bot., vol. xii, pp. 203-241, plates
xvii and. xvii, London, June, 1898,

23. Ikeno, $.—Untersuchungen tiber die Entwickelung der
Geschlechtsorgane und den Vorgang der Befruchtung bei Cycas
revoluta. Mit Tafeln vili-x und zwei Autotypien. Jahrbiicher fir
wissenschaftliche Botanik, Band xxxii, Heft 4, pp. 557-600,
Leipzig, 1898.

24. Worsdell, W. C.—The Gomync Anatomy of. certain
Genera of the Cycadacez. —Journ. Linn. Soc., London, July 1,
1898.

25, Marsh, O. C.—The Jurassic Formation on the Atlantic
Coast. Supplement. American Journal of Science, vol. vi, pp.
115-116, August, 1898.

26. Cycad Horizons in the Rocky Mountain Region.
Amer. Jour. Sci, vol. vi, p. 197, August, 1898.

27. Seward, nw C.—Fossil Plants, vol. 1, London, 1898.

28. Ward, ‘Lester F.—Descriptions of the Species of Cycade-
oidea, or Fossil Cycadean Trunks, thus far determined from the
Lower Cretaceous Rim of the Black Hills. Proceedings of the
U.S. National Museum, vol. xxi, pp. 195-229, Washington, 1898.

EXPLANATION OF PLATES.

Jeiwuviviop JOU

Cycadeoidea ingens, Ward (type).
Male plant; showing flower bud at summit.
. One-sixth natural size.

PLATE III,

Cycadeoidea ingens ; same specimen.
Longitudinal section through male flower.
For details, see diagram in text, p. 221, figure 1.
Twice natural size.

PLATE IV.

Cycadeoidea ingens; same specimen.
FIGURE 1.—Cross section through middle of male flower; showing sori and
bracts.
FIGURE 2.—Cross section through base.
For details. see diagram in text, p. 221, figure 2.
Both figures are twice natural size.

O. C. Marsh—Ffootprints of Jurassic Dinosaurs. 227

Arr. XXII.—Footprints of Jurassic Dinosaurs ; by O. .
: MarsH. (With Plate V.)

OnE of the most interesting geological discoveries during the
past season in the Black Hills region was a locality of foot-
prints evidently made by Dinosaurian reptiles in deposits of
Jurassic age. These footprints are the first found in Jurassic
strata in this country. They are all tridactyle, of large size,
and were evidently made by some of the great Dinosaurs
known to have lived during Jurassic time.

The tridactyle footprints hitherto found in this country were
nearly all discovered in the Triassic sandstone of the Connecti-
eut Valley, and, as well known, were at first supposed to have
been made by Birds. They have since been discovered in
essentially the same horizon in New Jersey and also in New
Mexico. In the same strata, many other footprints have been
found, and among them numerons similar tridactyle impres-
sions, some of which may possibly have been made by birds,
but by far the greater number are evidently of reptilian origin.
It is an interesting fact that the bones of Dinosaurs found in
this horizon of the Triassic all pertain to animals of moderate
size, and none are known large enough to have made any of
the gigantic footprints so abundant in the Connecticut Valley.

In the Jurassic formation of this country, on the contrary,
the osseous remains of Dinosaurs of large size are especially
abundant, and among these were not a few bipedal forms that
must have made footprints very similar to the so-called
bird tracks of the Triassic, but no footprints of any kind
have hitherto been found, although diligently sought for in
many Jocalities. The present discovery fortunately supplies
the much-desired information on this point, and the specimens
already secured promise to throw much light on the life-history
of this interesting group of reptiles that were the dominant
forms of life during Jurassic time.

When these footprints were first discovered, it was naturally
supposed they were of Triassic age, as all footprints of similar
character known in this country had been found in deposits of
that formation. The great development of red Triassic beds
in the same region, and their extension in a broad belt around
the Black Hills, where they are generally known as the “red
beds,” all seemed to favor such a supposition. The character of
these beds of red shale and sandstone, all evidently deposited
in shallow water, as shown-by the ripple marks and other well-
known features, indicated that footprints and other impressions
would certainly be found in them if proper search were made,
and this will doubtless prove to be the case.

228 O. OC. Marsh—Ffootprints of Jurassic Dinosaurs.

The large footprints here described, however, are from a
+higher horizon, and one well within the limits of characteristic
deposits of Jurassic age. Their position in the series of Meso-
zoic strata encircling the Black Hills is indicated in the dia-
gram, on page 229. Thissection is designed to show the succes-
sion of the principal geological horizons above the Paleozoic
in the Black Hills region. The lowest beds here represented
are the red beds, already mentioned, which are mainly fresh-
water deposits, and generally are referred to the Triassic
formation. :

Next above these comes a series of sandstones, limestones,
and shales, containing marine fossils, and named by the writer
the Baptanodon beds, from a genus of large swimming reptiles
there entombed. This horizon is readily recognized by charac-
teristic marine invertebrate fossils, especially Belemnites, first
described from this region about forty years ago by Meek,*
who recognized the importance of the horizon. Above these
marine beds are extensive fresh-water deposits of Jurassic age,
which the writer has called the Atlantosaurus beds, from a
gigantic Dinosaur specially characteristic of the horizon. It
is in this series of deposits on the southwestern border of
the Black Hulls that the footprints here described were
found, and it is a point of much interest that here, too, are
entombed remains of large reptiles that probably made the
same footprints, as will be shown later in the present article.

These Atlantosaurus beds, though overlooked by many geolo-
gists, have a great development around the margin of the Black
Hills, especially along the southern and eastern borders. The
bones of gigantic Dinosaurs mark the outcrop of this horizon
at various points. The one best known, the writer explored
personally in 1889, near Piedmont, South Dakota, and there
obtained remains of an enormous Dinosaur, subsequently
named Larosaurus.t During the past season, important parts
of the rest of the type skeleton were secured for the Yale
Museum, by G. R. Wieland of that University. With these
fossils were found remains of a much smaller species, which
may be called Barosaurus affinis.

Above the true Atlantosaurus beds is a series of strata of
shales and sandstones, the exact age of which is at present a
matter of controversy. In this series, there are two or three ~
layers which contain the remains of Dinosaurian reptiles
differing somewhat from those below, but especially from those
known in the well-defined Cretaceous strata above. The writer
has now under investigation for the U. 8. Geological Survey,
various remains of vertebrate fossils from one of these layers.
These fossils, with others secured from different localities in
the same region, promise to clear up many of the doubtful
points now remaining as to the age of these deposits.

* Proc. Acad, Nat. Sci. Phila., vol. x, pp. 41-59, 1859.
+ This Journal, vol. xxxix, p. 85, January, 1890.

0. C. Marsh—Footprints of Jurassic Dinosaurs. 229

The Cyead beds, as they may be termed, from the great
number and variety of remains of this group of fossil plants,
are abundantly represented around the rim of the Black
Hills, apparently at a higher level, but, as the Cycad remains,
although distinctive in themselves, have not yet been found

“absolutely in place in undisturbed strata, their exact position
in the series cannot at present be definitely fixed. Part of
this series was formerly referred to the Dakota by various
geologists, but this reference is fairly open to question, as the
writer has shown elsewhere.*

‘CENOZOIC.

- 1 Brontops, Allops, Titanops, Titano-
BrontotheriumBeds ? therium, Mesohippus, Ancodus, Entelodon.

|

Laramie Series. Nanomyops, Stagodon. Birds, Cimolopteryz.

Atlantochelys Beds
Cretaceous. lof Montana Group, Aantochelys, Coniornis.

MESOZOIC.

| Ctenacodon.

Pentacrinus.

{

| a
Triassic. Red Beds. |A few plants.

GEOLOGICAL HORIZONS ABOVE PALEOZOIC OF BLACK HILLS REGION.

Well-marked Cretaceous strata, showing the characteristic
yellow chalk of the Pteranodon beds of the Colorado series, are
developed east of the Black Hills, and contain an abundance
of characteristic fossils. These deposits the writer personally
explored in 1889, and proved their identity with the well-
known series along the Smoky Hill River in Kansas.

Next above come the Atlantochelys beds of the Montana
group, well developed, and marked by remains of gigantic
turtles, as well as by characteristic mollusea, and other inverte-
brates. The top of the Cretaceous east of the Black Hills is
formed mainly by this group. On the western side of the
Hills, the highest Mesozoic deposits are the Ceratops beds of the
Laramie. These form one of the best-marked horizons known
in any country, as here occur the gigantic horned Dinosaurs of
the genus Z?iceratops and others, as well as numerous small
Cretaceous mammals and birds.

* This Journal, vol. vi, pp. 107, 115, and 197, August, 1898. See also, the
present number, p. 219.

| Recent. ) Tapir, Pececary, Bison.
Quaternary. Bos, Equus, Tapirus, Dicotyles, Megatherium, Mylodon.
z | ; Equus, Tapirus, Elephas. .
_| Pliocene. Equus Beds § Pliohippus, Tapiravus, Mastodon, Procamelus,
P Phohippus Beds. 1 Aceratherium, Bos, Morotherium, Platygonus.
os |
Ss Protoceras, Miohippus, Diceratherium, Thinohyus.
= 'Protoceras Beds. § Oreodon, Eporeodon, Hyenodon, Ictops, Hyraco- |
=| Miocene. |Oreodon Beds. (don, Agriocherus, Colodon, Leptocherus.

‘Pteranodon Beds of Mosasaurs, Edestosaurus, Lestosaurus, Tylosaurus.
‘Colorado Series. Pterodactyls, Plesiosaurs, Vurtles.

\Cycads, Cycadeoidea.
Cy cad Beds. ‘Dinosaurs, Barosaurus, Brontosaurus, Morosaurus, |

2 | Diplodocus, Stegosaurus, Camptosaurus, Allo-
Jurassic. AtlantosaurusBeds. | saurus. Mammals, Dryolestes, Stylacodon, Tinodon, |

|Baptanodon Beds. 2aptanodon, Pantosaurus, Belemnites, Trigonia, |

Ceratops, Triceratops, Claosaurus, Ornithomimus. |
Ceratops Beds of Mammals, Cimolomys, Dipriodon, Selenacodon, |

|

|
J

230 O. C. Marsh—Ffootprints of Jurassic Dinosaurs.

The Eocene is apparently wanting in the Black Hills region, —
and the Miocene at many points rests directly on Cretaceous
strata. The Miocene is represented by three great horizons,
each marked by an abundance of mammalian remains. The
Brontotherium beds are at the base, the Oreodon beds next
higher, while the Protoceras beds crown the series. The Plio-
cene is also well developed, as shown in the section, with the
Pliohippus beds below and the Equus beds above. Over all,
the Quaternary and recent deposits are present, marked by an
interesting and characteristic fauna.

Below the series here given, the geological structure of the
Black Hills, as well known, is briefly as follows: The central
mass is Archean. Outside of this are Silurian strata with the
Potsdam sandstone at the base, and limestones of upper Silu-
rian age above. The Devonian is apparently wanting, and the
Paleozoic ends with Carboniferous strata, mainly limestones.

The age of the Atlantosaurus beds has long been demon-
strated to be upper Jurassic, by the conclusive evidence of
vertebrate fossils there entombed. Testimony from other
kinds of fossils has not been wanting, and many new facts are
coming to light. One noteworthy instance may be fitly
recorded here. Remains of the gigantic Dinosaur Barosaurus
are characteristic of a definite layer in the Atlantosaurus beds —
on the eastern side of the Black Hills. Just above this layer,
at various localities, there is a thin seam of arenaceous shale —
filled with remains of minute Ostracoda, and also containing a
few fishes. As this seam has a definite position, its fossil con-
tents became important in determining its exact age. Among
the fish remains, a perfect tooth was readily recognized as
fybodus polyprion, Ag., a characteristic Jurassic fossil, found
in the Dogger, at Stonesfield, England.

A small piece of this shale containing the ostracods was
sent to Prof. T. Rupert Jones of London, the highest author-
ity on the subject, and in a letter to the writer, dated February
9, 1899, he reports that in the specimen sent, three species —
were represented, namely, Cypridea punctata, Forbes, Jeta-
cypris Bradyi, Jones, M. Whiter, Jones, all characteristic of
the Purbeck. These new facts need no comment. It was
already known that two of these species, identified by the same —
author, occur in the same horizon of the Atlantosaurus beds in
Colorado, about four hundred miles further south.

The general characters of the large footprints here described
are well shown in those represented on Plate V, one-sixth
natural size. In figure 1, page 231, are given the outlines of
three other tracks, one-tenth natural size. These are from the
same locality and horizon as those on Plate V. They are all
reverse impressions, or natural casts, of the surface immedi-
ately over the true footprints, which were made in a soft mud
that has itself not been preserved sufficiently to retain the
imprints originally made in it.

O. C. Marsh—Footprints of Jurassic Dinosaurs. 231

The interpretation of the various footprints here shown is
fortunately a much less difficult matter than that of the huge
tridactyle tracks from the Triassic, as in the latter case there
was little except conjecture to assist in the investigation. In
the present instance, however, it is known that abundant osse-
ous remains of large Dinosaurs are imbedded in deposits of
the same age, and in the same general vicinity. Some of.
these reptiles are known to have been bipedal, and quite large
enough to make the footprints now discovered. Moreover,

the structure of the feet of some of these animals is sufficiently
well known to demonstrate clearly that in walking they must
have made footsteps similar to, if not identical with, the fossil

specimens here recorded.
1.

a. b. é

's,
S—aas=”

FiGuRE 1.—Outlines of Jurassic footprints found with those shown on Plate Y-
a, footprint of large herbivorous reptile; 0, footprint of large carnivorous
reptile; c, footprint of smaller carnivorous reptile.
All the figures are one-tenth natural size.

As to the particular reptiles that made the present foot-
prints, we have a hint in the impressions themselves. Some
of these show marks of short and wide toes, suggesting that
they were made by the robust feet of herbivorous Dinosaurs.
The other footprints show impressions of longer and more
slender toes, such as are known in the carnivorous Dinosaurs,
which lived at the same time and in the same region as the
herbivorous forms, and, indeed, preyed upon them. In Plate V,
the large figure on the left evidently represents the foot-
print of a herbivorous Dinosaur, while the small one on the
right, with slender toes, was probably made by a carnivorous
form. Of the outline figures above, the one with the thick
digital imprints (@) was doubtless made by a herbivorous rep-
tile, while the other two (d and c) appear carnivorous in type.

In selecting from this horizon the known Dinosaurs that
might have made the present footprints, it would be safe to
say that the genus Camptosaurus, especially some of its larger
species, may be held responsible for the herbivorous footsteps,
while its carnivorous enemy, Adlosaurus, had representatives
to which the more slender impressions may naturally be due.

232 O. OC. Marsh—Footprints of Jurassic Dinosaurs.

Other genera of carnivorous Dinosaurs are known to have
lived in Jurassic time in the same region where these footprints:
were made, and among them Creosaurus and Labrosaurus
would probably have left similar tracks, although the structure
of their feet is not accurately known. Ceratosaurus, another
carnivorous form, must also be considered, although not
known from the region where the footprints occur, while
Celurus is much too small to have made any of the impressions
yet discovered. Figures 2 and 3 below show the bones of
hind feet of Camptosaurus and Allosaurus, the Dinosaurs
above mentioned.

Figure 2.—Left hind foot of Camptosaurus dispar, Marsh.
Figure 3.—Left hind foot of Allosaurus fragilis, Marsh.
Both figures are one-twelfth natural size.

One point worthy of mention is, that both Camptosaurus
and several of the carnivorous forms are known to possess, in
addition to the three birdlike main digits,a rudimentary toe
on the inner side of the hind foot, representing the first digit.
This toe, however, was too short to make an impression in
ordinary walking on the ground, but might leave a mark
where the surface was very soft. No indications of such
maarks have been found with these footprints, however, and
could hardly be expected under the circumstances.

- The fossil footprints here described were found by H. F.
Wells, in 1898, in the Atlantosaurus beds, on the southwestern
border of the Black Hills in South Dakota. The specimens.
are now deposited in the Yale Museum. .

Yale University, New Haven, Conn.,

February 21, 1899.
EXPLANATION OF PLATE V. |
Footprints of Jurassic Dinosaurs from the Black Hills, South Dakota.
One-sixth natural size.

H.. L. Ward—New Kansas Meteorite. 233

Art. XXIV.—A new Kansas Meteorite; by Henry L.
WARD, Rochester, N. Y.

In October last Sam G. Sheaffer, Esq., Attorney-at-Law, of
Ness City, Kansas, called our attention to a meteorite that he
had in his possession ; and which, after some correspondence,
he sent to Ward’s Natural Science Establishment for the pur-
pose of disposing of it.

Mr. Sheaffer writes that “it was found about a year ago in
the southwest of this, Ness County,
Kansas. Was picked up on the side
of a draw, 1.e., a dry creek, where
the surface had been eroded.”

In form it is a triangular pyramid
with the base set obliquely to its per-
pendicular.

A mass of some weight had long
ago separated from the lower left
hand corner, as seen in the figure ;
but whether before, upon, or after
reaching the earth it is now impos-
sible to determine from the fractured
part. Several slight depressions
appear on the surface, which are
rather too sharply indicated in the accompanying figure. The
edges of the nearly plane faces meet in rounded angles.

The meteorite is 92™™ in length, 64™™ across its widest face,
left to right of figure, and 49™™ in thickness, measured _per-
pendicularly to the widest, and also largest, face. The termi-
nation had been chipped away for the purpose of ascertaining
its meteoric character before it was sent to us. The weight of
the mass is 417 grams.

This is not a prepossessing meteorite. It entirely lacks the

black crust characteristic of aerolites; and so strongly sug-
gested a weathered marcasite concretion that we were at first
skeptical as to its meteoric origin. However, tests for iron and
for nickel were both affirmative, and a polished chip showed
the former well distributed as minute specks through the mass.
A complete analysis has not yet been made.
_ To our knowledge but one other meteorite has been described
from Ness County, Kansas. That is the Kansada aerolite
designated by the name of the town near which it was found.
The locality whence came the specimen under consideration,
Section 2, Township 20, S. of Range 21 west, is not marked
by a town; and I therefore propose to designate this meteorite
as the Ness County.

234 - Scientific Intelligence.

SCIENTIFIC INTELLIGENCE.

I. CHEMISTRY AND PHysIcs.

1. On the Density and Composition of Liquid Air.—Experi-
ments have been made by LapENBuRG and KRUGEL on the density
of liquid air of varying composition. The method employed was
to weigh a glass rod, whose density at 4° was known, in the
liquid air contained in a silvered Dewar bulb, by means of a
Mohr balance, and to calculate the density from the loss in
weight. At the same time, the composition of the air was deter-
mined by Bunsen’s method, suitably modified for the purpose.
In this way three estimations were made. In the first, the
liquid air used was freshly prepared ; in the second it had been
allowed to stand some time until a portion of the nitrogen had
volatilized ; while in the third, it was not used until the second or
third day after its preparation. The density of sample one was
found to be 0:9951, and its content of oxygen 53°83 per cent;
that of sample two was 1°029, and its oxygen 64°2 per cent;
that of sample three was 1°112 and its oxygen 93°6 per cent.
Three facts appear here for the first time: first, that liquid air is
lighter than water; second, that the liquid air poorest in oxygen
contains over 50 per cent of this element; and third, that the
last portions, consisting of nearly pure oxygen, have a density
greater than that of liquid oxygen itself. Moreover the density
of the gas from this residue was found to be greater than that of
gaseous oxygen; it being 1°125 while that of pure oxygen is
11056. This is due probably to an admixture either of carbon
dioxide or possibly of krypton. From these data the density of
ordinary atmospheric air containing 20 per cent of oxygen, and
which is not yet obtainable in the liquid state as such, should be
calculable, either by geometric or algebraic methods. If the con-
tent of oxygen be plotted as abscissas and the density as ordi-
nates, three points are obtained lying nearly in a straight line.
Prolonging the line connecting either two of these points till it
cuts the ordinate drawn through 20°9 per cent oxygen, values for
the density are obtained lying between 0°887 and 0:908.. For air
free from oxygen, i. e., for pure nitrogen, a value of about 0°84 is
obtained, Olszewski having obtained 0°85. Or analytically, as-

suming that the three points lie on a curve of the second degree ”

such as y =a-+ ba + ca’, the constants a, b and ¢ can be calcu-

lated from the experimental values of y the density, and x the ‘

oxygen-content, above given. So that the equation becomes
y = 0'77892 + 0:00463x% — 0:00001152".
For the value x = 20°9 per cent of oxygen, this equation gives as

the value of y, the density, 0°8707. Hence were it possible to

liquefy atmospheric air without the more rapid evaporation of the
nitrogen, it would have a density between 0°87 and u'90. The
authors have applied the same method to the determination of the

—

Chemistry and Physics. 235

density of liquid oxygen and of ethylene. Both gases were liquefied
in U-tubes immersed in liquid air. Two experiments with differ
ent glass rods, gave for oxygen at the boiling point of liquid air (be-
tween —183° and —188°) a density of 1:105-1°108; these values
becoming 1°110 and 1°113 when the correction for the change in
volume of the glass is applied. Olszewski’s value at —181°4°, the
boiling point of oxygen, was somewhat higher, 1:110 to 1°137.
On passing ethylene gas into a tube immersed in liquid air in a
Dewar globe a crystalline mass of solid ethylene is at once ob-
tained, whose melting point was found to be —169°, and its
boiling point —105°4° at 760™™ pressure. The density of the
liquid ethylene at —169° was found to be 0°6585 and at —105°4°,
05710. Methane made from sodium acetate and barium oxide
gave variable results, probably from the presence of hydrogen
and perhaps of ethylene.— Ber. Berl. Chem. Ges., xxxii, 46, Jan-
uary, 1899. Goi.

2. On Semi-Permeable Membranes.—Experiments which have
been made by MisErs have led him to differ from the opinion of
Ostwald that semi-permeable membranes are not only conductors
like metals but are also “ion sieves.’’ He calls attention to the
fact that the copper which is deposited on semi-permeable mem-
branes possesses the nature of a spongy mass rather than a
coherent one. If a glass cylinder be closed by a membrane of
copper ferrocyanide on parchment. paper, and be filled with nor-
mal solution of copper sulphate, and then placed in an electrolyz-
ing vessel containing the same solution, the current from two
Bunsen cells being passed through it, it is found that no copper
is deposited on the membrane during the first half hour, a deposit
taking place only when the liquid in the cylinder has become
nearly colorless. On weighing the cathode, which is placed in
the cylinder closed by the membrane, both before and after the
experiment, it is observed that the amount of the deposited cop-
per is in excess, by a considerable amount, of that which was
previously contained in the solution in the cylinder, showing that
copper must have passed through the membrane. Placing a pla-
tinum cathode in acidulated water in a similar cylinder closed by |
a membrane, and putting this in an electrolytic vessel containing
copper sulphate solution, metallic copper is deposited in a few
minutes on the cathode; and after four hours the amount of this
copper is more than a hundred times that contained in the precip-
itated copper ferrocyanide on the membrane, which retains its
color and is not acted on by the acid. The author’s experiments
seem to show that semi-permeable membranes are conductors
differing entirely from metals. Thus if, in a copper sulphate
electrolyzing vessel the anode and cathode are separated by a
semi-permeable membrane held between two rings of ebonite, and
by a sheet of platinum of exactly the same size as the membrane,
no metal is deposited on the membrane when the current passes,
the deposit taking place on the platinum near the center, the
margin being protected by the ebonite.—Rev. Trav. Chim., xvii,
177, 1898; J. Chem. Soc., \xxiv (ii), 505, November, 1898.

G. Eu.

Am. Jour. Sc1.—Fourta Series, Vou. VII, No. 39.—Marcu, 1899.
16

236 Scientific, Intelligence.

3. On the Solution of Platinum and Gold in Electrolytes.—It
has been observed by Marcuetes that if an electro-magnet pro-
vided with a vibrating armature be made to act as a contact
breaker to the current from two Daniell cells, an electrolytic cell
provided with platinum or gold electrodes being included in the
circuit, the metal can be made to dissolve in hydrochloric, nitric
or sulphuric acid, or in potassium or sodium hydrate. The solu-
tion always takes place from that electrode which is for the time
being the anode. It is helped in some way, however, by the dis-
charge current from the magnet, as the current from the battery
alone does not effect the solution, neither when continuous nor
intermittent.— Wied. Ann., lxv, 629-634, June, 1898. rela ye 2

4. Aqueous Solutions of Metallic Gold.—By treating a feebly
alkaline boiling solution of gold chloride with a reducing agent,.
ZsicmMonpy has shown that a red aqueous solution of metallic
gold may be obtained.* Formaldehyde seems to be the best
reducing agent, less satisfactory results being given by acetalde-
hyde, alcohol and hydroxylamine. Sometimes the solution, in-
stead of being red, is dark purple, violet or bluish black and
appears turbid. The solutions of gold are very dilute, only about
0:005 gram of gold to the hundred cubic centimeters. They
undergo no change when boiled until reduced to less than half
their volume, when they become violet-black and deposit gold as.
a black powder. The solutions may be concentrated by dialysis,
which become intensely red and contain 0°12 per cent of colloidal
gold. Solutions of neutral salts and mineral acids change the
color to blue and slowly precipitate finely divided metallic gold.
Potassium ferrocyanide produces a green coloration, subsequently
becoming yellow. Acetic acid gives a violet red and finally a
black. Alkalies precipitate blue gold. Alcohol in excess changes
the color gradually to dark violet, precipitating the metal com-
pletely in a condition in which it is soluble in water. When sub-
mitted to the action of a current, aqueous solutions of gold act
like other colloids; the metal travels toward the cathode, but
does not penetrate the membrane.— Liebig’s Annalen, ccci, 29-54,
June, 1898. G. F. B.

5. On the Explosion of Mixtures of Methane and Air by the
Electric Current.—Hxtended experiments have been made by
Courtot and Mrvnizr on the action first of incandescent wires
and second of electric sparks, on mixtures of methane with air in
various proportions, the gas being sometimes at rest and some-
times in motion. A mixture of 80 per cent of methane and 20
per cent of air, in motion, could not be fired by an incandescent
wire nor even by the spark at the break when the wire fused ;
though it readily ignited in contact with a flame. If an electric
current produces any effect at all on a mixture of methane and
air, the only effect is an explosion. But under no circumstanc 2s
were the authors able to produce an explosion with an incandes-
cent wire. Whenever an explosion was produced it was due to
the spark produced at the breaking of the wire. A mixture con-

* See Faraday’s Exp. Res. in Chem. and Phys., p. 411.

Chemistry and Physics. 237

taining 9°5 per cent of methane is the most readily explosive.
The combustion is complete so long as the methane does not fal

below 5°5 per cent. Even with 4°5 per cent slight explosion

were noticed. With 12 per cent of methane, the explosibility
reaches its maximum limit, no explosion having ever been obtained
when the proportion was as high as 12°25 per cent.

In a subsequent paper, the authors state that if the current by
which the wire is heated be shunted with a parallel wire, the pro-
duction of the spark causes no ignition of the gas, if the resist-
ances are equal in the two branches. If the shunt-resistance is
high, or on the other hand if this resistance falls below a certain
value, the production of a spark always produces an explosion.
To avoid an explosion therefore the current strength must not
exceed a certain maximum depending on the resistance of the two
branches of the conductor.— C. F., cxxvi, 750, 901, March, 1898.

G. F. B.

6. On a Crystallized Compound of Cuprous Chloride and
Acetylene.—It has been pointed out by CaavasTELon that on
treating cuprous acetylide with hydrochloric acid in the cold, an
appreciable evolution of gas takes place ; due probably to the
decomposition on warming of a compound of cuprous chloride
and acetylene. Further investigation showed that this compound
may be prepared either by the action of acetylene on a solution
of cupric chloride in alcohol or in water, in presence of metallic
copper, or by passing this gas into a saturated solution of cuprous
chloride in dilute hydrochloric acid, at a temperature at or below
12°. By the latter method, which is preferable, large hexagonal
crystals are obtained belonging to the orthorhombic system, hav-
ing the composition C,H,: Cu,Cl,. By the former process, silky
crystals are yielded, which are liable to contamination by a violet-
purple deposit, formed early in the reaction. Both products alter
readily on exposure to the air, and are at once decomposed by
water or by solutions of alkali chlorides, with evolution of acety-
lene and the production of the above mentioned violet-purple sub-
stance. The crystals dissociate on warming without explosion,
producing at 78° a pressure of 262°".—C. #., exxvi, 1810-1812,
June, 1898. G. FAB.

7. Matter, Energy, Horce and Work, A plain presentation of
Fundamental Physical Concepts and of the Vortex-atom and
other Theories. By Siras W. Horman, Professor of Physics
(Emeritus) Mass. Institute of Technology. 12mo, pp. Xiv, 257.
New York, 1898 (The Macmillan Co.).—The aim of this book, as
the preface tells us, is to present some fundamental ideas and defi-
nitions of physics in a plain and logical manner. The subject
matter proper is contained in the first part, the definitions and
views there given constituting ‘a sporadic attempt at clear con-
secutive setting forth of individual thought.” The second part
consists of speculations on matter and energy. ‘wo new terms
are introduced, kinergety and weightal. Winergety in the abstract
sense denotes “the idea of capacity for kinetic energy” and in
the concrete sense “the quantity of this capacity.” Weightal is

238 Scientific Intelligence.

-defined to be “ quantity of substance measured by the equal-arm
balance.” With regard to the vortex-atom theory, Lord Kelvin
writes to the author “I am afraid it is not possible to explain all
the properties of matter by the vortex-atom theory alone.” “TI
have not found it helpful in respect to crystalline configurations,
or electrical, chemical or gravitational forces.” The book con-
tains a thoughtful presentation of fundamental physical concepts
and cannot fail to be of much service to those students who
desire a philosophic knowledge of modern theories. G. F. B.

8. Uranium Radiation, and the Electrical Conduction pro-.
duced by it.—A very exhaustive investigation of this subject has
been published by Prof. Rurwerrorp of McGill University.
Becquerel had stated that the uranium rays differed from the
Rontgen rays in this respect, that they can be refracted and polar-
ized. Professor Rutherford cannot find any evidence of refrac-
tion or polarization. The theory of ionization adopted by the
author supposes that the rays in passing through a gas produce
positively and negatively charged particles and that the number
produced per second depends on the intensity of the radiation
and the pressure. The term ion does not assume that the ion is
necessarily of atomic dimensions. These ions are supposed to be
carriers of electricity. Prof. Rutherford examines the theory that
energy 1s absorbed in producing the ions and that the absorption
is proportional to the number of ions produced and thus depends
on the pressure. If this theory is correct he points out that we
should obtain the following results :

(1) Charged carriers produced through the volumes of the gas.

(2) Ionization proportional to the intensity of the radiation and
the pressure.

(3) Absorption of radiation proportional to pressure.

(4) Existence of saturation current.

(5) Rate of recombination of the ions proportional to the square
of the number present.

(6) Partial separation of positive and negative ions.

(7) Disturbance of potential gradient under certain conditions
between two plates exposed to the radiation.

The experiments performed by the author indicate that the
theory affords a satisfactory explanation of the electrical con-
ductivity produced by uranium radiation. He shows also that
the uranium radiation is complex, and that there are present at
least two types, which he terms a- and f-radiations. The char-
acter of the B-radiation seems to be independent of the nature of
the filter through which it has passed, and passes through all
substances tried with far greater facility than the a radiations.
The photographic effects are due principally to the 6-radiations,
except when the uranium compounds are placed close to the pho-
tographic plate. The experiments on transparency of metallic
plates are not in accordance with those of Becquerel. The
absorption of the @-radiation in gases is probably of the same
order as that of the X-rays.—Phil. Mag., January, 1899, pp.
109-163. J. T.

Geology and Mineralogy. 239

9. On the Silver Voltameter and its use in determinations of
normal elements.—In a contribution from the Reichsanstalt, K.
Kaute discusses the subject and gives the results of a careful
inguiry of the electromotive force of a Clark and a Cadmium
cell, and a comparison with previously obtained results by means
of a Helmholtz electrodynamometer (Wied. Ann., lix, p. 532,
1896) of the electrochemical equivalent of silver. The final value
obtained is E = 11183 mg/sec. which is in accordance with the
previously obtained value.— Wied. Ann., No. 1, pp. 1-36, 1899.

agers

10. Absorption of light by a body placed in the magnetic field.—
Ruter discusses the work of investigators who have occupied
themselves with the phenomena of absorption discovered by him.
It is shown that the Zeeman effect is accompanied by a rotation
of vibrations of which the wave lengths are close to those of the
radiations absorbed. ‘According to Becquerel and Voigt, one
could predict this rotation from the facts of anomalous disper-
sion. Rhigi does not believe with MM. Corbino and Macaluso
that the rotation constitutes the principal cause of the appearance
of light in his experiments; for the condition that the phenom-
ena of rotation may be observable, namely, great length of rays of
absorption, is not necessary and is even prejudicial. One cannot
maintain absolutely that Rhigi’s experiments would lead to the
discovery of the Zeeman effect. They reveal the existence of a
phenomenon less simple—a Zeeman effect together with rota-
tions, more or less simple, of vibrations which, it may be, always
accompany it.— Comptes Rendus, No. 1, pp. 45-48, January, 1899.

IT

Il. Grotocy AND MINERALOGY.

1. Recent Earth Movement in the Great Lakes Region ; by
Grove Kari Gitpert. (From the Eighteenth Annual Report
of the Survey, 1896-97. Part II.)—The question as to the exist-
ence at the present time of progressive changes of level between
the water and land in the region of the Great Lakes is one not
only of great theoretic interest, but also of much practical impor-
tance, since a relative rise of the water may have a profound
effect upon the value of property on the shores of the Lakes.
This subject is discussed by Mr. Gilbert with great care and
thoroughness. He remarks upon the earth movements which
characterized the closing epochs of the Pleistocene Period, as
clearly brought out by the studies of Dr. Spencer, Mr. F. P. Tay-
lor, and others, of the shore lines of the early glacial lakes that
covered more or less of the region. He quotes Dr. Spencer (1894)
as urging that this change of level has not ceased, and that it will
eventually turn the water of the Upper Lakes southward to the
Illinois and Mississippi Rivers, leaving the Niagara channel dry.
This is one consideration which has weight in the matter.

Another is found in the condition of the estuaries which show

240 Scientific Intelligence.

a close resemblance to features observed along the subsiding
parts of the Atlantic coast, and give an impression that a slow
flooding of the stream valleys is still in progress. A third point
noted is the accepted fact that the Atlantic coast, south of Con-
necticut, is subsiding, as has been estimated (Cook) for New Jer-
sey about two feet inacentury; while the land about Hudson and
James bays has risen (Bell) perhaps five to seven feet in a cen-
tury. These facts all point in a common direction, warranting the
hypothesis that the tilting of the Lake region which went on at
the close of the glacial times, as shown by the slopes of old shore
lines, is still in progress.

The author notes at the outset a paper by Mr. G. R. Stuntz
(1869), in which he argued, in the case of Lake Superior, that
there was a gradual rise of water at the west end of the lake, and
a fall of the same at the east, referring this to a westward canting
of the basin, the western part becoming lower as compared with
the eastern.

With this introduction the author goes forward to discuss the
plan of investigation, and the data available, with also certain
special observations made in 1896. ‘The stations are taken in
pairs, the points being so chosen as to most satisfactorily show
what change of level, if any, has been going on. An absolute
conclusion is difficult to reach, since very few of the observations
are above question because of the doubt as to the stability of the
individual base levels involved. Still the candor and fairness
with which the author weighs the evidence is worthy of all praise
and gives his conclusions great value.

He decides that the harmony of the measurements, and their
agreement with the predictions of geological data, makes so
strong a case for the hypothesis of tilting that it should be
accepted as a fact, although some doubts exist concerning the
stability of the gauges. The mean rate of change deduced is
0:42 foot to one hundred miles in a century; but this depends
upon certain assumptions which he does not regard as altogether
probable, and he notes that the change indicated by Stuntz’s
observations is much more rapid. It would seem then that the
assumption is justified that the whole Lake region is being lifted
on one side or depressed on the other, so that its plane is bodily
canting toward the south-southwest, and that its rate of change
is such that the two ends of a line one hundred miles long, and
lying in a south-southwest direction, are relatively displaced four-
tenths of a foot in one hundred years. Certain general conse-
quences follow from this assumption (eliminating irregularities
due to difference of rainfall, evaporation, etc.). Thus, on Lake
Ontario, the water is advancing on all the shores. This is also
true of Lake Erie, the most rapid change, of eight or nine inches
in a century, being at Toledo and Sandusky. About Lake Huron
the water is falling more rapidly in the north and northeast. At_
Lake Superior the water is advancing on the American shore and
sinking on the Canadian. Similar relations exist for Lake Michi-

Geology and Mineralogy. 241

gan, the water falling in the north and rising in the south, at
Chicago some nine or ten inches. The ultimate result would be
the discharge of the lakes through the Illinois river, the channel
of the Pleistocene glacial lake being occupied anew. The high-
water lake discharge may begin in 500 to 600 years, for mean
lake stage it will begin in 1,000 and after 1,500 it will go on
uninterruptedly. In about 2,000 years the Illinois river and
Niagara will carry equal portions of the surplus water of the
4jreat Lakes, and in 3,500 there will be no Niagara. The author
adds : “'The most numerous economic bearings of this geographic
change pertain to engineering works, especially for the preserva-
tion of harbors and regulation of water levels. But the modifica-
tions thus produced are so slow as compared to the growing
demands of commerce for depth of water that they may have
small importance. It is a matter of greater moment that cities
and towns built on lowlands about Lakes Ontario, Erie, Michigan,
and Superior will sooner or later feel the encroachment of the
advancing water, and it is peculiarly unfortunate that Chicago,
the largest city on the lakes, stands on a sinking plain that is now
but little above the high-water level of Lake Michigan.”

The paper closes with a statement of certain definite plans
made for precise measurements at selected stations, for example, -
at Chicago, Port Huron, Parry Sound, and Mackinaw, which if
they can be carried out systematically, must in time lead to a
very definite solution of the problem in hand.

2. On the Petrology of Rockall Island.—-Professor J. W. Jupp
has recently described a remarkable rock under the name rockallite
from the isolated point which rises from a shoal above the waves
of the North Atlantic 240 miles west of Ireland. It appears to
be a remnant of an intrusive sheet resting on sediments. The
tiny islet has long been known under the name of Rockall. Prof.

Judd’s paper forms one of a series of memoirs descriptive of

this remarkable island.* As this rock is of very interesting char-
acter a brief description will be of interest to petrographers.

It is of granitic character but with an approach to a porphyritic
structure in places owing to the idiomorphic outlines of the lath-
shaped feldspars, which allies it to the granite porphyries. In
composition it is very simple, being made up of egirite, albite and
quartz in the following proportions: zgirite 39, quartz 38,
albite 23.

The egirite is found in larger crystals and in fine needles, and

shows at times brown acmite zones. The albite is twinned accord-

ing to the albite and Carlsbad laws and its phenocrystic sections
contain egirite. The quartz is the last mineral to torm and fills
the interspaces. A little apatite and soda amphibole are present.
The specific gravity varies from 2°9-2°7._ In the more porphyritic

* Notes on Rockall Island and Bank with an account of the Petrology of Rock-
all, and of its winds, currents, etc.: with reports on the Ornithology, the Inverte-

‘brate Fauna of the Bank and on its previous history. Transactions of the Royal
trish Academy, vol. xxxi, Part III, pp. 39-98, plates ix to xiv,

24.2 Sceventifie Intellagence.

variety a little groundmass is present. The chemical analysis
gave these results :

SiO. Al,0O3 Fe.,0; MnO NiO MgO CaO © Na.O K,0 P.O;
73°60 4°70 13°10 es ‘06 ‘11 “Nt 6°96 tr. tr. = '99°383

In this TiO, and FeO were not determined. The analysis is
remarkable for the low alumina, the high iron and the absence of
potash. In these respects it is most closely related to the groru-
dites of the south Norway region and to some of the lavas of
Pantellaria. From them it differs however in its granitic struc-
ture, the higher proportion of segirite and in the character of its
feldspar. The introduction of its name as a varietal one in the
alkaline granite group is therefore justified. Li yVieg Bs

3. Lecherches géologiques et pétrographiques sur le massif du
Mont Blane ; by L. Duparc et L. Mrazec (Mem. Soe. de Phys.
et Hist. Nat. de Genéve, xxxiu, No. 1, 47) ppsg22 tees.
1898).—In this great work the first part is devoted to a full
account of the topography of the Mont Blane massif and the
position it occupies in the system of the Alps. Then follows a
very full petrographic investigation of the eruptive rocks and of
the relations existing between them followed by a similar account
of the crystalline schists which accompany them. A very large
number of analyses of the types described is also given.

The fourth part is devoted principally to the description and
theory of the metamorphic phenomena produced by the proto-
gine in the crystalline schists with which it is in contact. The next
part is a description of the sedimentary rocks involved in the
mass and the whole concludes with a discussion of the tectonic
processes and of the folds and uplift which they have produced.
In addition to the large number of half-tone plates of scenery
and rock sections, a number of excellent geologic sections in
color are given. ‘The authors regret that the expense prohibited
an accompanying publication of the geologic map they have made,
a regret in which all readers will agree with them.

The whole is an exhaustive monograph and gives the results of
an enormous amount of painstaking labor both in the field and in
the laboratory. It is one of the most important contributions to
Alpine geology which has appeared in recent years. Ts ayViggl

4, Native Silver in North Carolina; by G. F. Kunz. (Com-
municated.)—In June, 1898, while some mining was being carried
on at the West Prussian Mining Company’s land, at Silver Hill,
near Livingston, Davidson Co., N. C., an interesting deposit of
native silver was brought to light. An inclined shaft was sunk
to a depth of 75 ft., near a mixed mass of sulphides of lead, zinc,
copper and iron, associated with a vein of green rock, in places
weathered and decomposed almost to a white clay. In and
through this, a soft, slaty mass of native silver was found, dis-
seminated in grains and plates presenting the appearance of hay-
ing been absorbed in the interstices of the slate, and varying in
size from minute grains and scales of metal to pieces several

Geology and Mineralogy. 243

square inches in area, and up to one-fourth of an inch in thick-
ness; the largest single mass weighing, with some of the adherent
slate, about 5 lbs.

The sulphides are not crystallized; but appear to be almost
homogeneous and are comparatively rich in lead and zinc, an
average assay yielding 35 per cent of zinc and 20 per cent of
lead, with 25 ounces of silver and one-fourth an ounce of gold to
the ton. An assay made by Dr. A. R. Ledoux proves the silver
to be almost absolutely pure, containing only 1 per cent of native
gold. The silver occurs in veins, with more or less open spaces,
and showing very minute crystals on the upper and lower part of
the vein. A near view, side-ways, reveals it to be in fibrous
masses and ropes, having all the appearance of sweating out of
the rock, probably a result of the decomposition of the chloride,
leaving the silver in fine, fibrous masses, varying from short ropes
10 millimeters in diameter to the fineness of a hair.

The specimens and information were kindly furnished by
Wyndham H. Wynne, of Glendalough, Ireland.

5. On the occurrence of polycrase in Canada; by G. Cur.
HorrMann. (Communicated.)—Fine examples, of what on ex-
amination by Mr. R. A. A. Johnston proved to be polycrase, have
been found by Mr. C. W. Willimott, in the township of Calvin,
district of Nipissing, in the province of Ontario. Jt occurs here
in the form of crystalline masses—one of which weighed rather
more than seven hundred grams—associated with xenotime, a
highly altered, cleavable massive form of magnetite, and small
quantities of a brownish-red spessartite, in a coarse granite vein,
composed of quartz, microcline, albite or oligoclase, muscovite
and biotite, which is there found cutting a reddish, fine-grained,
hornblendic gneiss. The mineral has a pitch-black colour; an
uneven, in parts subconchoidal fracture; a resinous lustre; is
brittle; and affords a grayish-brown streak. Its specific gravity,
at 15°5° C., is 4°842. A very carefully conducted qualitative
analysis showed it to contain—Niobic oxide, large amount; tan-
talic oxide, somewhat small amount; titanic oxide, large amount;
yttrium oxide, somewhat small amount; thorium oxide, small
amount; stannic oxide, trace; cerous oxide, small amount; lan-
thanum oxide, small amount; didymium oxide, small amount ;
uranous oxide, small amount; ferrous oxide, small amount ; mag-
nesia, trace; water, very small amount. Zirconia was sought for,
and found to be absent.

6. The Kaolins and Fire Clays of Europe, and the Clay-
working Industry of the United States in 1897; by HErnricu
Ries. From the Nineteenth Annual Report of the U.S. Geologi-
cal Survey, 1897-1898.—Professor Ries has here brought together
a large amount of important information on the occurrence, com-
position and general use of the clays from the important localities
in Europe. ‘This is chiefly based upon the personal observations
of the author during recent visits to the localities themselves.
The value of the paper will be appreciated when the importance

244 Scientific Intelligence.

of the industry in this country is considered, and at the same time
the very imperfect and fragmentary amount of information which
has already been published in the various technical journals. A
brief summary is also given of the clay industry in the United
States.

7. On the Origin of the Gases evolved on heated Mineral Sub-
stances, Meteorites, etc.—A paper by M. W. Travzrs in the
Proceedings of the Royal Society (No. 405) discusses the evolu-
tion of gases from various minerals and meteorites, especially
with reference to their origin. His conclusions are somewhat at
variance with those ordinarily accepted, in that the experiments
go to prove that in the majority of cases, the gas evolved under
the influence of heat is the product of decomposition or inter-
action of non-gaseous constituents present in the substance under
examination. It is conceded that compact minerals do inclose
carbon dioxide and hydrocarbons as easily liquefiable gases; but
the analogy cannot be extended to gases such as hydrogen and
helium in connection with readily cleavable minerals like chlorite,
mica and cleyeite. The special importance of the conclusions here
reached lies in the fact that it invalidates the conclusion reached
in regard to the origin and history of the substances in question.
‘This is partly true in the case of meteorites.

IIJ. Botany AnD ZooLoGcy.

1. Symbolee Antillane seu Fundamenta Flore Indic Occiden-
talis ; editit Ianarius Ursan. Vol. I, Fase. I continet: I, Ign.
Urban: Bibliographia Indiz occidentalis botanica, pp. 3-192.
Berolini, 1898 (Fratres Borntraeger).— Within recent years, Pro-
fessor Urban has published numerous papers dealing with the
flora of the West Indies. His articles have appeared in various
German journals and have of necessity been somewhat scattered
and disconnected. It is now his intention to publish the further
results of his studies in this domain in a series of papers under
the above title, the various fasciculi to appear at indefinite periods
and to deal mainly with difficult or neglected families, the descrip-
tion of new genera and species and the geographical distribution
of plants on the islands. This first paper contains a list, arranged
alphabetically according to authors, of the numerous works deal-
ing with the botany of ‘the West Indies, but it is much more than
a mere enumeration. After each title, which is given in full and
also in a shortened form suitable for citation, follows an abstract
of its contents, and, in the case of taxonomic papers, an indica-
tion as to where the plants described are at present preserved.
Especially valuable are the many references to works published
in Spain or in the West Indies themselves, many of which are
difficult of access and therefore little known to botanists.

A. W. E.
Plant life: considered with special reference to form and
pattie by Cuartes Reip Barnes, Professor of Plant Physi-

~

Botany and Zoology. 245

ology in the University of Chicago; pp. x +428, with 415 figures
in text. New York, 1898 (Henry Holt & Company).—As stated
in the preface, Professor Barnes’ text-book attempts “to exhibit
the variety and progressive complexity of the vegetative body ;
to discuss the more important functions; to explain the unity of
plan in both the structure and action of the reproductive organs ;
and finally to give an outline of the more striking ways in which
plants adapt themselves to the world about them.” The four
parts into which it is divided deal accordingly with the following
subjects: I. The Plant Body; II. Physiology; III. Reproduc-

tion; and IV. Ecology. It will be seen, therefore, that the book

is very comprehensive in its plan and that most of the great
divisions of modern botany receive a certain amount of attention.
The topics are, with few exceptions, treated according to recent
researches, but the very comprehensiveness of the subject makes
great conciseness necessary, and it is to be feared that some of
the statements are so concise as to be obscure, at least to pupils
from thirteen to eighteen years of age, for whom the book is pri-
marily intended. The arrangement of the subject-matter, also,
although logical, might readily lead the pupil to believe that a
plant is composed of certain definite and more or less discon-
nected parts, instead of being an organism whose parts, although
distinct enough from a formal standpoint, are nevertheless so
intimately connected and so dependent upon one another that
they form a most definite whole: in other words, that botany is
the study of plant-ergans rather than of plants. For example,
under “‘ simple sporangia” on page 316, the sporangium of Mucor,
the ascus of Peziza and the tetrasperangium of Polysipbonia are
described in due order; but, if we wish to find other facts about
Polysiphonia, we must look on page 32, for a description and
figures of the vegetative structure, and on page 289, for a figure of
the cystocarp: and other, even more striking examples might be
quoted. The short section on ecology, which, as the author
implies, is largely abridged from two recent German works, is an
interesting feature of the book, and, it is to be hoped, will demon-
strate to teachers how very useful this important department of
botany might become for purposes of instruction. At the close

of the book are five appendixes in which laboratory directions,

lists of apparatus, reagents and reference books, and a very brief
outline of classification are given. A. W. E.

3. Rhodora: Journal of the New England Botanical Club,
Vol. I, Nos. 1 and 2, January and February, 1899. Boston and
Providence.—After mature consideration, the New England
Botanical Club has decided to publish a journal devoted to the
interests of New England botany. The first two numbers have
recently appeared, and the editorial announcement states that
the journal will devote special attention to.matters connected .
with the geographical distribution of New England plants and to
the revision of difficult or misunderstood groups of species. It
also states that “not only flowering plants, but ferns, mosses and

246 Scientific Intelligence.

thallophytes will receive their proportionate share of attention.”
Some idea of the character of the matter presented may be
obtained from the following titles, selected from the numbers
before us: rattlesnake plantains of New England (with plate);
notes on aleze; a new wild lettuce from Massachusetts (with plate);
notes on some fleshy fungi found near Boston; some plants about
Williamstown ; fairy-rings formed by Lycopodium; bryophyte
flora of Maine. A. W. E.

4, Catalogue of the Lepidoptera Phalene in the British
Museum, Vol. I, Syntomide; by Grorcr F. Hampson. With
17 colored plates. 8vo. London, 1898.—This is the first volume
of a proposed catalogue of all the known genera and species of
moths. The introduction contains a brief statement of the
external structure of the Lepidoptera and an analytical key to
the families. This is followed by the account of the Syntomide,
a family of very pretty moths which isconfined almost exclusively
to tropical and subtropical regions. Nearly twelve hundred
species are described and the genera and subgenera illustrated by
285 figures in the text. The plates contain about 370 beautifully
colored illustrations of new or hitherto inadequately figured
species, but, as they are not essential to the plan of the work,
they are issued separately, so that the text may be sold at a much
less price than if the plates were combined with it.

TV. MiIscELLANEOUS SCIENTIFIC INTELLIGENCE.

1. Report of S. P. Langley, Secretary of the Smithsonian Insti-
tution, for the year ending June 30th, 1598.—The annual reports of
the working of the Smithsonian Institution are always of much
interest, because of its remarkably wide range of activity. One
of its most important functions is the international exchange ser-
vice, and the map of the world here introduced, which shows the
distribution of its correspondents, sets forth in a very telling man-
ner the perfection to which this system has been carried. Not
only are correspondents very numerous in the well-known coun-
tries, but we find, for example, 99 in China, 204 in Brazil, 24 in
Iceland, etc. The total number of correspondents is very nearly
30,000.

Much of interest is given in regard to the workings of the
Smithsonian in every direction, but we can only speak particu-
larly of the Astrophysical Observatory. Mr. C. G. Abbott, in
charge, enumerates as follows the most important features of
the work of the past year:

1. The instrumental equipment has received valuable accessions,
including a highly sensitive galvanometer, designed and con-
structed at the Observatory ; two cylindric mirrors by Brashear
(which, as used for collimation of the spectroscope, are equivalent
to a lens of 64 meters focal length), and, finally and most import-
ant of all, a system of cooling by the expansion of ammonia,
which has made possible an extension of constant temperature

Miscellaneous Intelligence. 247

conditions to cover the five months of March, April, May, Sep-
tember, and October, otherwise frequently too warm. At pres-
ent the change of temperature of the inner room between these
warmer and the coldest winter months is only a fraction of a
degree centigrade, and during an hour’s observation it is gener-
ally less than one-tenth degree, the control being automatic.

2. Many bolographs of the infra-red solar spectrum have been
taken, which, in consequence of these improvements, have
yielded results threefold richer in “real” detail corresponding to
solar and telluric absorption lines than any hitherto obtained.

3. About 40 of these bolographs have been compared, as de-
scribed in the Appendix to the Secretary’s Report for 1896, and 21
of the most perfect have been measured upon the comparator to
determine the positions of the deflections found to be “ real,” or,
in other words, corresponding to either solar ortelluric absorption
lines. These comparator measurements included about 44,000
separate observations.

There have thus been found over 700 absorption lines in the
infra-red solar spectrum between wave-lengths 0°76 yu and 6°0 y,
an increase of about 500 over last year’s results.

4. With the purpose of making a more accurate determination
of the wave-lengths corresponding to the well-determined posi-
tions of the absorption lines discovered in the rock-salt prismatic
spectrum, a very exact comparison of the dispersion of rock
salt and fluorite has been made. This comparison will allow the
indirect employment of certain recent and apparently very accu-
rate determinations of the wave-lengths in the fluorite prismatic
spectrum. Apparatus has been made ready and certain prelimi-
nary observations have been taken to directly measure the dis-
persion of rock salt. It is hoped that these steps will result in
furnishing the wave-lengths of the infra-red absorption lines to a
degree of accuracy corresponding to the exactness of the deter-
mination of their prismatic deviations.

5. Many interesting instances of local variations in the absorp-
tion have been noticed. Among these by far the most striking 1s
a great decrease in the absorption at the longer wave-length side
of the great band 7 at about 1°4 yw. This change occurred about
February 15, 1898, and caused the bolographs to take on quite a
different form at the place in question. This new form continued
through the months of March and April, but in the month of
May the usual form was gradually restored. It is found, by
reference to former bolographs, that this marked decrease in
absorption at this point takes place annually at about the same
period, which coincides (fortuitously or otherwise) very nearly
with that at which there is the greatest activity of growth in the
vegetable kingdom. This raises the question whether the growth
of vegetation does not abstract from the air great quantities of
some selectively absorbing vapor active in absorption at this wave-
length.

Whether such be the case or not future investigation must

248 Scientific 1 ntellagence.

determine, but enough variations in the absorption have beem
observed to indicate that the Observatory is now in condition to
make advances along the line indicated in the Secretary’s Report.
for 1892, in which is pointed out the important relations of astro-
physics to meteorology.

2. Physical Geography ; by Wit1am Morris Davis; pp.
XVill, 428, with nine plates, etc. Boston, 1899 (Ginn & Co.).—
This volume has the necessary elements of a high grade text-
book: it is scientifically accurate, contains no unnecessary matter
and is written ina pleasing manner. The earth’s features are not.
only described, they are explained, and the reader is led to see a
rational meaning in the varied land-forms about him. The
student is constantly reminded, also, of the way in which phys-
ical features and climate have influenced man in his settlements,
industries and manner of life.

The illustrations deserve special praise both in their selection
and mechanical execution, The numerous “ block-diagrams ” are
particularly valuable. Education in general and the out-of-door
sciences in particular will be benefited by the introduction of just
such text-books into secondary schools. H. E. G.

3. The Annual Report of the Director of the Field Columbian
Museum to the Board of Trustees, for 1897-98. Report Series,
Volume I, No. 4. Chicago, October, 1898.—This Report gives an
interesting summary of the work of the Field Columbian Museum
for 1897 and 1898. The numerous illustrations introduced give
views of many of the rooms with their collections, and also of
some of the objects of particular importance.

OBITUARY.

Prof. Louis W. Pox, for several years of the department of
Physics in the University of Minnesota, died in Tucson, Arizona, on
December 26th, in the 47th year of his age. He was born in Provi-
dence, R.I1., and was educated in the Mass. Institute of Technology.
He also served anapprenticeship in the Harris Corliss Engine Works
in Providence. Prof. Peck’s contributions to science were made in
the investigation of the flouring mill explosion that occurred in
Minneapolis, Minn., May 2, 1878. ‘The mills exploded and then
burned. They were insured against loss by fire and it became a ques-
tion of great importance whether fire caused the explosion or the
explosion caused the fire. Prof. Peck contrived a simple appara-
tus by which he showed that flour and wheat dust, when mixed
with air in suitable proportions, would explode with terrific force
if ignited with fame. Glowing charcoal and white hot platinum
wires would not ignite the mixture, hence the conclusion was
inevitable, confirmed by collateral evidence, that the mills were
on fire before they exploded. Prof. Peck’s results were embodied
in a paper which was published in the Popular Science Monthly,
Xlv, p. 159, and the London Journal of Science, xvi, p. 666. This
paper was widely copied and reviewed. Bud IPs

“Two Shintaae from Colorado bring us a rich assort- on ee
ment of rare tellurium minerals, all of which have been

most carefully tested and identified, our labeling being 5.
guaranteed. Calaverite, 25c., 35¢., $1.50 and $2.00. Sag
Altaite, $2.25. Coloradoite, $5.00. Petzite, 25c. Oe
to $3.00 ; one extra fine large specimen, $12.50. Hessite, a

25¢. to $1.00. Sylvanite, 35c. to $2.56. Tellurium, “HR
10c. to $4.00, a few of the specimens well crystallized.
Colorado Gothite, in groups of radiated crystals,

occasionally terminated, 25¢ to $1.25. a
Minium from Colorado,"excessively rare, 75¢c, $3.00 | . par ‘
and $6.00. eG ee

Martite from Utah, groups of mammoth crystals, 25c. to $3.50.

Selenite from Ohio; 2vU finest quality two-inch crystals, 20¢. each; very se,

good crystals, 5c. to 15c. aS

_ Canadian Titanite crystals, 975 of them at 5e to Tic. each. ions
Fs Aragonite erystals from Bohemia; 100° choice loose crystals at 15c. to 50¢.

‘Japanese Quartz Twins. As previously announced, we have secured all
the very choicest of these wonderful specimens, probably the most
eresting specimens from a scientific standpoint which have been in the Pree
arket during the present decade. All but one of the finest of the large i aaa
ins haye been sold at prices ranging from $30 to $82.62. One fine = js Ree
iirix specimen with an exccllent four-and-one-half-inch twin, $65, is in stock as oe
2gotopress. We also have a small assortment of good. but more bruised. ees
ins ranging from 14 to 34 inches, $2.50 to $12.50. As these twins are not a wee
sew find, as none of them have been found for several years notwithstanding dili- ee
ge nt search has been made for them, and as we have had all of the best (except gti
one ‘sold to a California college), our customers can feel perfectly safe in purchas- Sas
ing now, as no more good ones are known in Japan and only a few poor ones, ieee
ected for the most part by us, are on sale outside of our store.

aad other most desirable minerals have been recently secured.

GEO. lh. ENGLISH & CO., Mineralogists,
| REMOVED TO
312 aud 814 Greenwich Street (S. W. Corner of Jane Street), New York City. 3

\ CHOICE COLLECTION OF METEORITES
FOR SALE. =

The undersigned offers for sale for $250 his private collection of meteorites. .
e collection embraces 50 falls, aggregating 1,184 grammes, each specimen ys,
eled by Dr. Aristides Brezina, late of the Royal Museum, Vienna, and the.
ognized authority: on meteorites. The specimens are mostly flat pieces, and
" were carefullyrselected for the owner by Dr. Brezina so as to form a thoroughly
_ Tepresentative study collection. It is rarely that so excellent yet so inexpensive
a collection is obtainable. Pingo! the falls are: Alfianello, Italy: Bear Creek,
Colorado; Bridgewater, N. C.; Cosby’s Creek, Tennessee; Cross Timbers, Texas ;
agle, Kentucky ; Fort Picrre, Missouri; Juirnas, France; Kenvale Co., Texas; Pe
Kernouve, France: Misteca, Mexico; Nelson Co, Kentucky ; Ochansk, Russia ;
Paci, Mexico: Seelasgen. Prussia ; ‘Silver Crown, Wyoming; Summit, ‘Alabama,
A. complete list of the specimens with weight of each will be mailed on request.

GEORGE L. ENGLISH, |
812 Greenwich St., New York City. 1

CONTENTS:

Arr. XVII.—Studies in the Cyperacee ; by T. Hotar___-
XVIII.—Granitic Breccias of yi, Peak, Col.; : by Gs He

XJX.—Constitution of the Ammonium- Ns atioat Phos- “B
phate of Analysis; by F. A. Gooca and M. Austry ._

XX.—Crystal Symmetry of the Minerals of the Mica Gree s.
by T. L. Waikerie

DOL: —Descriptions of imperfectly known and new Actinians,
with critical notes on other species, IV ; re A: EB. Vero

XXII.—Study of some American Fossil Cree Part i;
The Male Flower of Cycadeoidea ; by Go ie Wasa
(With Plates II-IV.)

XXIUI—Footprints of Jurassic Dinosaurs ; 5 byeO: C. Maren.
(With Plate V.)

XXIV.—New Kansas Meteorite ; by H. ly Wanp

SCIENTIFIC INTELLIGENCE,

Chemistry and: Physics—Density and Composition of Liquid Air, Taneeaes
and KRUGEL, 234.—Semi-Permeable Membranes, Miyers, 235. ~Solution of
Platinum and Gold in Electrolytes, MARGUELES: Aqueous Solutions of Metallic —
Gold, ZsigmonDY: Explosion of Mixtures of Methane and Air by the Electri
Current, Courtor and MEUNIER, 236.—Crystallized Compound of Cuprou
Chloride and Acetylene, CHAVASTELON: Matter, Energy, Force and=Work, &

_W. Hoiman, 237.—Uranium Radiation, and the Electrical Conduction produce
by it, RUTHERFORD, 238. —Silver Voltameter and its use in determinations o
normal élements, K. KAHLE! Absorption of light by a body placed-igthe mé

netic field, Rurer, 239. :

Coa and Mineralogy—Recent Karth Movement in the Great Lakes ae
G. K. GILBERT, 239 .—Petrology of Rockall Island, J. W. Jupp, 241.—Recher

ches eéologiques et pétrographiques sur Je massif da Mont Blane, L. Dupart
and L. Mrazec: Native Silver in North Carolina, G. F. Kunz, 242.—Occurrenee
of polycrase in Canada, G. C. HOFFMANN: Kaolins and Fire Clays of Europe
and the Clay-working Industry of the United States in 1897, H. Rigs, 243,-
Origin of the Gases evolved on heating Mineral Substauces, Meteorites, ete.
W. TRAVERS, 244. &:.

Botany and Zoology—Symbole Antillane seu Spee eae Figres Indie Oceid
talis, I. URBAN: Plant life, considered with special reference to form an
function, C. R. BARNES, 244,—Rhodora: Journal of the New England Botani- |
cal Club, 245 —Catalogue of the Lepidoptera Phaleenz in the Briti h Muset mM; |
Vol, I, Syntomidze, G. F. Hampson, 246. a

Miscellaneous Scientific Intelligence—Report of S. P. Langley, Secretary o of
Smithsonian Institution, for the year ending June 30th, 1898, 246.— Physic
Geography, W. M. Davis: Annual Report of the Director of the Hild. z
bian Museum, 248. ‘

Obituary—Louis W. PxEcK, 248.

D. Walcott,

U.S. Geol. Survey. ee

Sit ae > Ape (hid
So's es ; » er
Ba ck Pipe us 1
N hates De
Patt iat 5 (Re, it
j Le ae
“
. “
, ee
1899
°
-

APRIL,

i

\

_ Established by BENJAMIN SILLIMAN in 1818.

» AMERICAN
OURNAL OF SCIENCE.

Hpitors: EDWARD S8. DANA.

ASSOCIATE EDITORS

Prorzssors A. E. VERRILL ann HENRY S. WILLIAMS,
or New Haven,

Beason GEORGE F. BARKER, or Puimapztpuias,
Prorrssor H. A. ROWLAND, or Batruorg,
Mr. J. S. DILLER, or Wasurneron.

FOURTH SERIES.
VOL. VII—[WHOLE NUMBER, CLVII.]

No. 40.—APRIL, 1899.

WITH PLATES VI-VIL.

NEW HAVEN, CONNECTICUT.
1899.

TUTTLE, MOREHOUSE & TAYLOR, PRINTERS, 125 TEMPLE STREET.

*

Published monthly. Six dollars per year (postage prepaid). $6.40 to
subscribers of countries in the Postal Union. Remittances should
© either by money Orders; rearstered letters, or bank checks.

Low Prices on Showy Minerals. —

During the past month of stock-taking, we have brought to light many large
De Satan oe and attractive case specimens, which have been occupying space needful for the
oe accommodation of laree consignments expected. It is necessary to secure quick |
sales, and of course the price governs this important point. The list following
does little to convey the idea of the liberal values offered. Note also that many
of these minerals are among our ‘‘leaders,” prices of which were unusually low,
and now when that figure is halved, or quartered, the nonen ee of the oppor
tunity may be appreciated.
Specimens are intended for open case exhibition, and sppromaneie sizes. vary
between 4 by 5 and 10 by 15 in.—making displays which are sure to attract
deserving attention. The steadily increasing size of our immense stock necessi-
tates this sacrifice, and it will be to the readers’ advantage if one or more oe
these BLOTS, are ordered promptly. ‘

Sulphur, crystallized, 5 by 6, 75c. to $1.50; 8 by 12, $1.00 to $2.50. Sicily.
Chalcopyrite, bright tetrahedrons on Dolomite. Size varying between
6. by 5 and 8 by 12, 50c. to $1.50. - Joplin, Mo..
Pyrite, from various localities, 4 by 6 to 8 by 12, 50c. to $1.50.
Marcasite, in large specimens from Missouri. Low indeed at 50c. to $2. 00.
Hematite, finely “crystallized and some iridescent, from Elba, 75c. to $1.50.
Sphalerite, rough crystals, Missouri, 75c. to $2.00, in 6 by. 'g groups.
Turquoisin veins Masses 4 by 6, 50c. to $1.00; 6 by 10, $1.00 and $2.00.
-Fluorite, England and N. Y. Groups 5 by 6, 7 Se, to $1. 25: ; larger up to $2. 50.
Calcite, England, Missouri and Mexico. ae 50c. to $1. Ov for 4 by. 6,
larger for $2. 00.
Stalactites up to 15 in. in length, 35c. to $1. 25, pre eae iy a
Selenite, Utah, clear cleavage and crystals, 75c. to $2.00. ores
Selenite, Sicily, fish-tail twins, cheap at 75c. to $2.00.
Pearl Spar, Mo. Specimens 4 by 6, 85c. to 75c.; 8 by 12, g1.0 00 to $1. 50.
Barite, bright groups from England, 75c. to $2.00.
Thenardite, crystal aggregates 4 by 6, $1.00 to $2.50; 8 by 12 and ~ over,
$3.06 to $5.00. Formerly rare. .
Hanksite, a few 6 by 8 and over, $3.00 to $6.00.
Calcite after Hanksite, Colorado, 4 by 6, T5c. to $1.25. .
Colemanite, large crystallizations, 75c. to 83, 00.
Quartz, Rock Crystal. Attractive groups from Hot Springs, 75ce. to 99; 00.
Quartz, Jasperized Wood. Masses showing bark, 75c. to $1.50. .
Opalized Wood. Masses retaining the wood structure, THC. 40% BICC
Amazon Stone. An immense stock. Specimens 4 py 6, -d0ch fo S150 -.
larger up to $2.00. . . Rf
Labradorite. Gem material. Large masses, 35c. to $1.00. - sate
Garnet, large dodecahedron in schist, Colorado. 5c. to $2.00.
Rubellite. Masses 4 by 6, 50c. to 75c.; 6 by 8 and over, $1.00 to $3.50.
Verd-Antique, Polished. Sections 12 by 12 and ae $2.00.

The above will be sent on approval at customers’ expense, no charge made for — 9
packing. If any specimens are not retained, their return should be prepaid.
All orders will have prompt attention and should be sent in at once to

Dr. A. E. FOOTE,

WARREN M. FOOTE, Manager.

1317 Arch Street, Philadelphia, Pa., U.S. i

OTHNIEL CHARLES MARSH

PROFESSOR OF PALEONTOLOGY IN YALE UNIVERSITY,
VERTEBRATE PALEONTOLOGIST TO THE UNITED STATES
GEOLOGICAL SURVEY, AND FOR MANY YEARS AN
ASSOCIATE EDITOR OF THIS JOURNAL, DIED AT HIS
HOME IN NEW HAVEN AFTER A BRIEF ILLNESS, ON

MARCH 18, IN THE 68TH YEAR OF HIS AGE.

a oe gh

THE

[FOURTH SERIES. ]

| Ant. XXV.— Glacial Lakes V ewberry, Warren and Dana, in
Central New York ; by H. L. Farrcuinp. (With Plate VI.)

CONTENTS:

Page. Page.
eee eo ln «249 | Dake Warren... 22 2 cL 257
_ Description of the Maps _-.._----- 250 General description ______._- 257
-Physiography of the district..-.__ 252 Invasion of New York __.__- 258
_ Earlier local glacial lakes _____._- 253 Shorelines and elevations -.-. 258
ee Newberry. 2... _-_._---- 255 | DDS GC) HOO a ee ee 259
Meme ee 255 | Hyper-Iroquois waters .---....--_ 260
Extent and tributaries -_---.-- 255 Lake Dana (“Geneva” beach) 260

Shorelines and elevations- -... - 256 Water levels between Lakes
Siretmcwon’..22 2.522222 - . 257 Dana and Iroquois _._----- 262
. Takeninegimis serach ok 263

Introduction.

THE ice sheet of the last glacial epoch covered, during its
greatest extent, all the area of the Great Lakes. When the
_ receding front of the waning glacier had passed to northward
of the southern boundary of the Laurentian basin, the glacial
and meteoric waters were impounded between the ice front
and the north-sloping land surface. These glacial lakes had
their outlets southward across the divide, and they expanded
- northward as the barrier of ice receded. ‘The local lakes were
gradually united by draining of higher levels to lower, and
ultimately these glacial waters became of vast extent, exceed-
_ ing in breadth and depth any existing lakes. Their shorelines
have been found at many localities and traced for great dis-
_ tances and the romantic history and relationship of the glacial
waters have become in recent years the subject of a special
body of earth-science literature.

Am. Jour. Scr.—Fourts Series, Vou. VII, No. 40.—AprRiL, 1899.
17

250 A. L. Fairchild—Glacial Lakes in Central New York.

During the later phase of the Laurentian glacial waters,
Lake Warren invaded central New York from the west, and
drew down to its level the local waters of the Finger Lakes
valleys. This region is of special interest, on account of its
singular topographic character, and because it includes the
critical locality where the great Warren waters found lower
escape to the sea (than by the westward outlet to the Mississippi)
and excavated in their flow toward the Mohawk a series of
remarkable channels, piling the debris in great deltas in the
more quiet waters of the intersected valleys.

The abandoned rock channels and correlating deltas are the
most conspicuous featares produced by the glacial waters.
Southeast of Syracuse these great cuts in rock, some of them
headed by cataracts, lie so close together that they cannot be
represented satisfactorily upon the map (Plate VI). They
are the last records of a lake history of truly dramatic interest.
Their examination will repay not only the special student in
geology and physiography, but the person having merely a
general or casual interest in those sciences.

The study of the glacial lake phenomena in New York State
has now been carried so far that the principal facts and rela-
tions in the history of those waters are known. It is the
purpose of this writing to describe briefly the succession of
events in the life and extinction of the later and broader glacial
waters in the critical district of the Finger Lakes.

Description of the Maps, p. 251 and Plate VL.

The larger map locates in a generalized way the shorelines of
the four greater glacial lakes, Newberry, Warren, Dana, and
Troquois.

The outlet of Lake Newberry was south to the Chemung
River through the col (900 feet altitude) at the head of the
Seneca Valley. The probable eastward limit of this water is
indicated between Cayuga and Owasco Valleys, and the locality
of its destructive drainage or lowering into Lake Warren is
shown east of the north end of Canandaigua Lake.

The Warren waters found their first eastward escape south-
east of Syracuse by either the channel west of Butternut
Valley, with altitude 840 feet, or the one east of that valley
with elevation 890 feet. (The channels are better shown on
the smaller map.) The uncertainty as to the earliest spillway
is indicated by the broken termination of the eastern end of
the Warren shoreline. The most westerly of the hypo-Warren
outlets is the one north of Skaneateles with elevation 812 feet.
East of this channel is a series of channels which conveyed the
overflow eastward to lower and lower levels. |

251

E
iy
=
=
MN
sS
s
S
Ss
-§
%
ro
s
N
SS
=
S
Ss
i
as)
os
aS
S
NS
‘Ss
&
Nj
Ry

868]
QA as QRS EE at tit

JOVNIVYOQ JMVT IVIOVTO

$v1130 Jo

SiNVNW3Y Sidvyvivd STINNWHD

hajje, Ayn fo}

}yip1eD ©

=
gamer LT

y

Dnt

f
ISNIVYAS "

te LL
» ga
ae

SSE SEASSSSSSSSSS

s)
*

|

RaSSSS

Y)

Se
pnysor ©

OUILPAPN ©

(

169

snjja210WW©
099

952 HA. L. Fairchild— Glacial Lakes in Central New York.

Lake Dana is a stage of the waters having their level deter-
mined by the outlet channel east of Marcellus, elevation 691
feet, with which the Dana shoreline terminates. |

The several rock channels cut by the east-flowing waters in.
their escape to the Mohawk-Hudson lhe so near together that
they cannot be well represented on the larger map (Plate VI),
but they are more clearly shown on the smaller map, p. 251.

The small circles on the larger map indicate cities or villages,
the names of-which are omitted to save confusion. On the
smaller map the names are given. |

The heavy broken line indicates the post-glacial divide
between Laurentian and south-flowing waters. Bars drawn
transverse to the water-parting indicate the location of outlets
of the primitive local glacial lakes; those elsewhere on the
map represent later outlets of secondary waters.

Numerals give elevations, in feet, above ocean level.

Physiography of the District.

The territory in general which includes the glacial lake phe-
nomena described in this paper is that part of New York State
known as the region of the “ Finger” Lakes. Its more pre-
cise limitations are, on the west, the valley of Honeoye Lake
and on the east a line somewhat beyond Syracuse. Those
north and south valleys lying to the west of Honeoye, namely,
Canadice, Conesus and the Genesee, were never related to the
Newberry waters.

The accompanying map shows the hydrography of the present
time. The topography will be sufficiently described, with the
help of elevations given on the map, by stating that the area
between Lake Ontario and the northern ends of the Finger
Lakes is a plain of 400 to 600 feet elevation, while the north
and south ‘ Finger” Lake and intermediate valleys are pre-
glacial trenches in an elevated tract of more enduring rocks.
As far as the waning ice sheet upon the north left the land
exposed, the northern and lower plain was covered by the
Warren and Dana waters, which extended southward in aseries
of deep narrow bays up the valleys into the higher plateau.
To the work of these lakes, with the subsequent Iroquois, is
due the heavy silt deposits and the subdued topography of the
lower part of this plain.

The ice movement, in the closing phase, followed the direc-
tion of the lake valleys, thus having a radial or spreading How ;
and local lobes of the ice, or stream glaciers, protruded beyond
the general ice front up each of the valleys, leaving in each
valley a heavy moraine, which now constitutes the head of the
present drainage, or the divide between north and south flow.

HA. L. Fairchild—Glacial Lakes in Central New York. 253

The ice was probably thicker in the deeper and broader valleys
of Seneca and Cayuga, and the general direction of the retreat-
ing front of the great ice body was probably convex toward
the south, and for our theoretical discussion may be regarded as
normal to the axes of the divergent valleys.

Since the glacial lakes were drained the beaches have been
thrown out of horizontal position by a differential northward
uplift that has affected the entire Laurentian area. It is there-
fore evident that during the life of the glacial lakes the alti-
tude of Central New York was much less than now, and that
the depression was progressively greater toward the north or
somewhat é¢ast of north.

Earlier Local Glacial Lakes.

As the front of the ice sheet melted back upon the north of
the divide, the larger valleys in Central New York were occu-
pied by important local lakes, which had their overflow across
the divide into southern drainage. These lakes have been
described with some detail in former papers by the writer.*
For the present purpose it will be sufficient to enumerate the
lakes and briefly state their location and relationship.

The glacial Honeoye occupied the valley of that name and
sent its overflow into the Canandaigua Valley. At an earlier
time the Honeoye Valley was flooded by the Canandaigua
waters (Naples Lake).

The Naples Lake was the earliest glacial water in the Canan-
daigua Valley, with its outlet across the divide at the head of
the valley, near Atlanta station. Subsequently a lower escape
was found across the eastern border of the basin into the Flint
Creek Valley and the, waters became tributary to those in the
Keuka Valley (Hammondsport Lake). This later phase of the
Canandaigua waters is called the Naples-Middlesex Lake.

The Italy Lake occupied the upper Flint Creek Valley and
overflowed at the head of the valley into Naples Lake. A
lower stage of Flint Creek Valley waters, the Potter Lake,
has not been fully studied at the time of this writing.

The Hammondsport Lake occupied the Keuka Valley. This
valley had a lake history similar to that of the Canandaigua.
The early phase of the Hammondsport Lake had its outlet
south through Bath village. A later phase had its outlet
across the eastern rim through the village of Wayne, but ulti-
mately to the Cohocton Creek, like the earlier phase.

* Glacial Lakes of Western New York, Bull. Geol. Soc. Amer., vol. vi, p. 353 ;
Genesee Glacial Lakes, ibid., vol. vii, p. 423; Lake Warren Shorelines in Western
New York and the Geneva Beach, ibid., vol. viii, p. 269; Glacial Waters in the
Finger Lakes Region of New York, ibid., vol. x, p. 27-68.

954 A. L. Fairchild—Glacial Lakes in Central New ¥ork.

The Watkins Lake occupied the upper Seneca Valley with
outlet at Horseheads to the Chemung River. This was the
predecessor of Lake Newberry, the latter being simply the
former lake widely expanded into the Cayuga and other valleys
adjacent to the Seneca Valley. (See page 255).

The Ithaca Lake occupied the Cayuga Valley with an outlet
southeast to Catatonk Creek and the Susquehanna River.

The Groton and Moravia Lakes occupied the Owasco Valley,
the former being the primitive lake in the main valley with
outlet at Freeville, the latter being the later and larger lake
with lower outlet southwest; both stages being tributary to-
the Cayuga waters. .

The first glacial Skaneateles, in the valley of that name,
overflowed by the head of the valley past Scott village and
Homer to the Tionghnioga Creek, but a later and much lower
stage, called the Mandana Lake (from the name of its outlet).
received the overflow of glacial waters of the Otisco and
Onondaga Valleys, and had its outlet westward into the
Warren waters, which then occupied the Owasco Valley.

The glacial Otisco overflowed at first southward past Preble
village to the Tioughnioga Creek, but a later stage of the
waters, the Marietta Lake, had outlet westward into the lower
stage of the Skaneateles. At the foot of this valley, and also
of the Onondaga and Butternut Valleys, we find the great
channels and deltas, produced by the eastward flow of Warren
and later waters, which are to be discussed in this paper.

The Cardiff Lake occupied the upper Onondaga Valley and
overflowed through the site of the Tully Lakes to the Tiough-
nioga Creek. <A later, lower lake in the middle section of the
Onondaga, the south Onondaga Lake, had its outlet by two
successive channels westward to the Marietta Lake in the
. Otisco Valley.

Butternut Valley was long occupied by a primitive glacial
lake, called the Butternut Lake, with outlet near Apulia sta-
tion to the Tioughnioga Creek below Tully village. When
the valley was. nearly clear of ice the glacial waters escaped
for a time westward into the Onondaga Valley.

Eastward is still another north and south valley involved in
this history, that of Limestone Creek. The earlier local lake
phenomena have not been studied, but the lower (northward).
part of the valley was occupied by the hyper-Iroquois waters
and related to the great eastward-leading river channels.

In the eastern valleys, Skaneateles, Otisco, and Onondaga,
the later glacial lake history is intricate and with surprising
changes of drainage. It will be noted that the later, local
glacial waters drained westward—those of the Butternut
Valley into the Onondaga, the Onondaga into the Otisco, the

H. L. Fairchild—Glacial Lakes in Central New York. 255

latter into the Skaneateles and this into the Owasco Valley,
then occupied by the great Lake Warren. The westward-leading
channels of these earlier waters are shown on page 251. When,
however, Lake Warren found eastward escape, all these valleys
were’ drained eastward with the lowering Warren or hyper-
Iroquois waters.

Lake Newberry.

Origin.—The lowest of all the cols across the line of water-
parting separating the present north and south drainage in New
York is the one at Horseheads, the head of the Seneca Valley,
with elevation of 900 feet. This great channel* is the final
result of the erosion by the concentrated glacial drainage from
the district lying between and including Honeoye Valley on
the west and Owasco Valley on the east. Originally the chan-
nel carried only the overflow of the Watkins Lake and was |

doubtless somewhat higher than now, but the amount of the

down-cutting has not been determined. The name “ Watkins”
lake applies to the waters as long as they were limited to the
Seneca Valley. Lake Newberry may be regarded as having
its birth when, by the recession of the ice front, the waters of
either the Ithaca or the Hammondsport Lake were drained
down to and became confluent with the Seneca Valley waters.
Lake Newberry is therefore only the larger, more expanded
Watkins Lake. The former name is used, not only as a matter
of convenience but also for uniformity in glacial lake nomen-
clature ; in order to distinguish the broad waters with complex
relationship from the smaller, more localized,-and simpler
Watkins Lake.

Judging from the topography of the region, it seems more
probable that the first local lake to lose itself i the Horseheads
level was the Hammondsport. The locality of its overflow was
probably near 2d Milo, four miles south of Penn Yan.

Fixtent and tributaries—Upon the west, Keuka Valley
became a gulf of the Newberry waters, first by the Penn
Yan depression and later by the open valley to the north.
Flint Creek Valley became a bay of the greater lake when the
ice receded from the locality of Stanley and Gorham. The
waters of Canandaigua and Honeove Valleys were certainly
tributary to the Hammondsport Lake through the Flint
Valley, and probably continued tributary to the Flint Valley
waters even after they were a part of Lake Newberry.

Upon the east, the Ithaca Lake found its close when the ice
withdrew as far north as Ovid. The scourways made by the
escaping Ithaca waters show an east and west erosion of the

*The Horseheads channel is described and figured in a former paper, Bull.
Geol. Soc. Amer., vol. vi, p. 365.

256 H. L. Fairchild—Glacial Lakes in Central New York.

drumlin surface in and south of Ovid village. In the Cayuga
Valley the Newberry waters existed for a time sufficient to
produce clear evidences of its presence, but as to Owasco
Valley the evidence is negative. They may have existed there
a short time, or the ice dam toward the Canandaigua Valley
may have lifted and drained Lake N ewberry before the removal

of that which barred the Owasco waters. The earlier Owasco

waters (Moravia Lake) were certainly tributary to Lake New-
berry by the North Lansing-Ludlowville channel and probably
were the easternmost of such tributary waters. :

Lake Newberry certainly included the valleys of Flint Creek, .

Keuka, Seneca, and Cayuga Lakes, and possibly Owasco fora — |

brief time. In form the lake was an east and west belt of
water resting against the ice front, and with four deep bays
reaching southward up the valleys above named. Its linear
extent was considerable, and its shoreline great, but its total
area was not large in proportion.

Shorelines and Elevations.—The phenomena of the New-
berry shorelines have been observed, chiefly as terraces, and
wave-built bars on terraces, of landstream deltas, throughout
the Keuka and Seneca Valleys and in the southern part of the
Cayuga Valley. These have all been described in detail in
another writing* and it will be sufficient here to briefly men-
tion the localities and altitudes.

In the Seneca Valley the Newberry shore phenomena have
been measured at many localities; at the head of the valley
between Odessa and Montour Falls (Havana), 935 to 940 feet ;
at Montour Falls, 950 feet (aneroid). On the west side, at
Watkins, 960 feet ; at Himrods, 990 (¢) feet (aneroid); at
Stanley 970 feet. Between Reed’s Corners and Ennerdale
occur the channels which drained the lake to the Warren level,
985 feet to 930 feet (aneroid). On the east side of the valley
are good delta ridges at Burdett, 964 feet ; at Hector’s Station,
976 feet ; at North Hector Station, 984 feet ; at Lodi, 977 feet.
At Ovid the level is in the village, but the exact elevation is
undetermined.

In the Cayuga Valley well-defined beach ridges occur at
Trumansburg, 962 feet elevation. No effort has been made to
locate the shore phenomena at other points, but they can
undoubtedly be found throughout the valley.

In Keuka Valley the main delta terraces throughout the
valley are 955 to 970 feet elevation, and are attributed to New-
berry waters.

The northward rise of these altitudes is evident and meas-
urable, but the matter of beach deformation will not. be
discussed in this paper.

* Glacial Waters in the Finger Lakes Region of New York, Bull. Geol. Soc.
Amer., vol. x, p. 27-68.

ia

Re
‘®
=

H. L. Fairchild— Glacial Lakes in Central New York. 257

Fxtinction.—In the study of the glacial lake history, it was
recognized that Lake Newberry must have been destroyed by
outflow either to the east or west, eastward by the Mohawk
Valley, or westward into Lake Warren. it was a matter of.
the quicker removal of the ice barrier in one or the other of
two critical localities. And wherever the escape occurred we

should find the scourways across the north-sloping land, or

some transverse configuration with traces of erosion by the

_ escaping waters, and with a favoring relation of valleys or

openings beyond. With some knowledge of the topography
of the critical east and west localities and their general condi-
tions and relations, it was thought more probable that the
Newberry waters escaped westward, and an examination of the
eastern area at the foot of Skaneateles and Otisco Valleys, sup-
plemented by a study of the Skaneateles sheet (unpublished)
ef the New York topographic map, confirms this view of the
Newberry extinction. The writer has found no evidences of
water at the Newberry level in the Skaneateles Valley, but

_ such should occur if the waters found egress eastward, because

the low Mandana pass (900 feet), east side of the middle of the
valley, would have let in the waters long before the ice dam
was removed from the high ground northeast of the valley.
Indeed no conclusive evidence has been found of Newberry
waters in the Owasco Valley, west of the Skaneateles. More-
over, the channel cut by the eastward escape of the subsequent
and lower Warren waters lies only a few miles north of the
high ground at the north end of Skaneateles Valley, and
within this short distance of critical ground there are no east

_and west scourways or any evidence of water-flow between the |

Newberry and the Warren levels.

Lake Newberry was destroyed by draining westward over a
point of land about five miles southeast of Canandaigua village,
and midway between Ennerdale railroad station and Reed’s
Corners (see Plate VI). Here are well-defined and capacious
channels, at successively lower levels, lying athwart the ice-
moulded or drumlin surface. They occur on the south side of
a morainal series and truncate the ends of drumlins. Here
the Newberry waters were lowered to the Warren level, being
drained from about 980 feet elevation down to 880 feet, and at
the lower level the waters of the former Newberry area became
a part of the great Lake Warren.

Lake Warren.

General Description.*—This glacial lake was of great area.
It is believed to have covered at least the southern part of the
Huron basin, all of the Erie basin, and as much of the Ontario

* A fuller description will be found in Bull. Geol. Soc. Amer., vol. viii, p. 269.

958 HT. L. Fairchild—Glacial Lakes in Central New York.

basin as was not still buried under the great ice-body. The
beaches have been traced in the Huron-Erie basin and along
the southern slope of the Erie and Ontario basins into Western
New York as far east as the meridian of Rochester. The out-
let of the lake was certainly westward into Mississippi drain-
age, probably across Michigan into the glacial lake Chicago.
The name was given by J. W. Spencer.

Invasion of Central New York.— The Warren waters
existed in the extreme western part of the state, within the
Erie basin, for a long time before they were allowed to enter
the central part of the state. When the front of the ice bar-
rier moved back, northward, from the high ground north of
Batavia, Lake Warren extended eastward, as a relatively narrow
belt of water, along the east and west ice-wall, with southward
prolongations in the valleys. The waters occupied the valley
of Oatka Creek to beyond Warsaw ; of the Genesee and Cana-
seraga Valleys to beyond Dansville;- the valley of Conesus, and
ultimately the valleys of Honeoye, Mud Creek (Bristol), Canan-
daigua, I'lint Creek, Seneca, Keuka (through Seneca), Cayuga,
Owasco, Skaneateles, Otisco, Onondaga, and possibly Butternut.
The valleys of Hemlock and Canadice were above the Warren
level.

The breadth and southward extent of Warren waters in the
valleys of Keuka, Seneca, and Cayuga were less than that of
the Newberry waters, as they were about 100 feet lower. The
northern boundary, the ice front, was probably much straighter
or less lobate than that of Newberry. The glacial border
being further north, on the low northern plain, the effect of —
the Finger Lakes valleys was nearly or entirely lost upon the
ice-front, which was probably an arcuate line not far from the
north ends of the present lakes.

Shorelines and Elevations—The Warren shore has been
traced with sufficient continuity to positively identify it as far
east as Lima, south of Rochester, where it has an elevation of
879 feet. Well-defined beaches may doubtless be found
further east, but the shoreline phenomena already discovered
are delta terraces and bars and spits connected with deltas.
Like the Newberry phenomena, these have been elsewhere
described and need only be mentioned here.

In the Canandaigua Valley the Warren level appears upon
the deltas, but the elevations have not been measured. A
fragment of beach has been found southeast of Canandaigua
village and one mile west of Reeds corners with elevation of
about 880 feet. In the Keuka Valley the Warren level is very
evident as the lowest of the strong delta terraces. These vary
in elevation from 830 to 870 feet. In the Seneca Valley sev-
eral terrace bars have been measured as follows: At Montour

Et. L. Fairchild— Glacial Lakes in Central New York. 259

Falls, 840 feet; between Montour Falls and Odessa, 840 feet ;
at Himrods, 853 feet (?); at Hector Station, 872 feet; at Lodi

_ Station, 874 feet; at Ovid, 877 feet.

In the Cayuga Valley the Warren levels have been measured

_at Coy Glen, 831 feet, with a higher bench 865 feet ; at Willow

Creek, 829 to 863 feet ; at Taughannock Falls, 826 to 849 feet ;

at Trumansburg, 829 to 834 feet. In the Owasco Valley, at

Locke, 860 to 870 feet, and at Moravia, 855 to 882 feet (ane-

_ roid). The two levels in the Cayuga and Owasco Valleys are
_ significant. These vary from 23 to 34 feet apart and are
thought to represent the fall from the normal Warren level

down to the first enduring level with eastward discharge, possi-
bly the rock sill at the head of the channel southeast of James-
ville (see pages 251 and 260). Th® lower level is hypo- Warren,

or, more accurately, hyper-Iroquois.

Fixtinction.—The life of Lake Warren was terminated by

_ the opening of a new and lower outlet, eastward, past the ice-

front, to the Mohawk-Hudson. For a long period, perhaps

several thousand years, its waters had been tributary to the

aa

Mississippi, but for some centuries they had been creeping
eastward along the ice front as if in patient search for a new
escape that was sure to come. It should be kept in mind that
the district we are considering has been uplifted to its present
level since that time; it then lay at a much lower altitude.

The most western spillway of the eastern discharge of
Warren waters is on the crest of the Corniferous escarpment
four miles west-northwest of Marcellus village, and the same
distance north-northeast of Skaneateles. The present elevation
of the sill or intake at the lowest point is given by Mr. Gilbert
as 812 feet, which is perhaps 80 or 90 feet under the theoreti-
cal Warren plane. There has been some removal of overlying
drift and possibly of shale, and the intake has probably
migrated backward to the present point. But this outlet, the
“Gulf,” was probably not the earliest nor highest spillway.
In another writing (see reference, p. 256) reasons have been
given for believing that the Warren waters at their full height
invaded the region as far east as the Onondaga Valley and pos-
sibly the Butternut Valley. Their first and highest eastern escape
was east of the Onondaga Valley, in the vicinity of James-
ville (see map, p. 251) and the waters were quietly lowered
upon the hollows between drumlins at the head of the Gulf
channel.

For years after the eastern escape was opened Lake Warren
must have had two outlets and divided its contribution of
surplus water between the Atlantic and the Gulf of Mexico ;
but soon after the Gulf outlet was opened it doubtless
carried the entire overflow. Whenever this occurred the lake
was no longer Lake Warren.

260 H. L. Fairchild—Glacial Lakes in Central New York.

This large channel, excavated in Hamilton shales by the
Warren overflow, leads southeast four miles, opening into the
Otisco Valley two miles south of Marcellus village. The local
name is the “Gulf,” which, for want of a better name is
retained. The waters were again impounded in the Otisco
Valley, where the debris derived from the excavation of the
Gulf was piled in a delta, with terraces from 860 down to 800
feet elevation. The Otisco waters (Marcellus Lake) escaped
by another great rock channel, Cedarvale channel (named after
a hamlet in the gorge) that heads at Marcellus village and leads
southeast to the Onondaga Valley at South Onondaga.

These great channels, with others farther east (see page 251),
show the final result of some centuries of erosion. Their
initiation was probably not in the present geographic sequence,
and the primitive conditions of elevation, etc., are uncertain. -
The Skaneateles and Tully sheets of the New York State topo-
graphic map will, when published, show the Gulf and Cedar-
vale channels and deltas with clearness.* Dr. Gilbert was the
first to study these cliannels and to point out their relationship
to those lying further east.f

HHyper-Iroquois Waters.

Lake Dana (“ Geneva” beach).—At the time of Warren
extinction the Ontario basin was still largely occupied by the
glacier. Certainly most of the eastern part of the basin north
of the locality now under discussion was filled with the ice.
Considerable time was required for the front of this ice-body
to recede so far as to permit an expanse of Ontario waters in
the region under discussion, with immediate-outlet at Rome—
that is to say, for the westward expansion of Lake Iroquois.
During this interval the glacial waters fell about 500 feet, from
the Warren plane to the Iroquois. This subsidence was not a
steady or continuous lowering, but per saltwm,—a series of
comparatively sudden falls, as progressively lower outlets were
opened, with intervals of steady level, or very slow falling, as
each outlet remained effective.

During one of these phases of repose it is believed that the
beach was formed which has been traced along the west side of
the Seneca Valley and described in a former paper as the
“Geneva” beach.t It is also believed that the level of the
beach correlates with the Cedarvale outlet at Marcellus. This
is the only waterlevel between the Warren and Iroquois planes

* To the U.S. Geological Survey the writer is indebted for advance copies of
the maps in manuscript which have facilitated this study.

+ Old Tracks of Erian Drainage in Western New York, by G. K. Gilbert. Bull.
Geol. Soc. Amer., vol. viii, p. 285.

t Bull. Geol. Soc. Amer., vol. viii, p. 281; vol. x, p. 44.

| A airchild-— Glacial Lakes in Central New York. 261

that has left any strong shoreline phenomena, so far as at
present observed. Theoretically its evidences should be found
from Marcellus westward throughout the Ontario and Erie
basins, and it may be the subject of much future study. A
distinctive name is required, and it is proposed to call the
water Lake Dana, after Professor James D. Dana, the early
leader in American glacial geology.
Since the publication of the former description the beach has
been traced south from Geneva to near Dresden, and shoreline
phenomena have been observed in the Cayuga Valley at a
theoretically corresponding level. The elevations in the
_ Seneca Valley are as follows: near Angus station, Fall Brook
yailroad, 671 feet; near Earle’s station, 679 feet; west of
_ Geneva, 700 feet; south of Phelps, 713 feet. The deforma-
_ tion of the beach is about three feet per mile, on north and south
line, and it lies 180 feet under the Warren plane. Evidences
of this lake should be found throughout the valleys of Cayuga
_and Seneca and westward at 170 to 190 feet under the Warren
level. This water could not occupy the valleys of Keuka,
Owasco, Skaneateles, nor Otisco, as these lie above its plane.

The magnificent Cedarvale channel, with its huge delta, has
_ been described in another writing (see reference, p. 256). It
heads on Corniferous limestone and shale, one-fourth mile
southeast of Marcellus village, with an average elevation of
691 feet, which elevation correlates well with the beach alti-
_tude of the same parallel, something over 700 feet, making
_ allowance for depth of water over the sill and the later down-
cutting, with some east and west deformation of the land.

The change from the Warren level to the Dana level did not
_ come by the slow down-cutting of the head of the Gulf chan-
nel, but by the recession of the ice front so that the glacial
waters abandoned that outlet and found free access to the
Marcellus valley from the north. .

The eastern limit of Lake Dana was practically the same as
that of Lake Warren, but with a surface 180+ feet lower. Its
southern margin must have been more contracted, while the
ice boundary upon the north was somewhat further removed.
Its western limits have not been determined, but phenomena
attributed to this water have been observed as far west as the
Genesee River. Without some unexpected barrier, like great
land elevation or far readvance of the ice sheet, there seems no
escape from the conclusion that Lake Dana must have covered
the Erie basin. Being weak as compared with the Warren
phenomena, these have been overlooked, but it is predicted that
they will be found westward over a wide area. The hypo-
thetical shoreline is indicated on the map.

262 Ff. L. Fairchild—Glacial Lakes in Central New York. —

Water levels between Lakes Dana and TIroquois.—In the —
district south and east of Syracuse there are several other
abandoned channels. All are lower than those described
above; some of them are in rock and of great capacity. The
most interesting ones are the sluiceways which carried east- —
ward the overflow of Warren and Dana Lakes, and which con-
stitute, with the Gulf and Cedarvale channels, a connected
series (see page 251). All this series lie transverse to the
present drainage and conveyed eastward, across the intervening
high lands, the waters impounded by the ice-dam in the north
and south valleys. The flood poured into the Onondaga —
Valley from the hypo-Warren and Dana Lakes by the Cedar-
vale channel was carried over to the Butternut Valley by three
successive channels lying southeast of Syracuse and near James-
ville. The lowest and most important is a great “rock cut,”
three miles southeast of Syracuse. While perhaps not the
largest, this channel is the finest, the most typical, and the
most easily seen of any glacial channel in New York. It is
about three miles long and is traversed, its whole length, by —
the Delaware, Lackawanna & Western railroad, and has been
named the ‘ Railroad” channel. The walls are nearly vertical
bare rock, 125 to 150 feet high, and the bottom of the cut will
average 900 feet wide.

The water was conveyed eastward from the Butternut
Valley to the Limestone Valley by other great rock channels,
the largest and latest one heading opposite the mouth of the
Railroad channel and extending somewhat over three miles to
High bridge, south of Fayetteville. The Syracuse sheet of
the topographic map well portrays the Railroad channel, but
the High Bridge channel is not well delineated on the north
side by the 20-feet contours. From the valley of Limestone
Creek three channels lead northeast ; the highest of these heads
in a cataract, one and one-half miles southwest of Mycenae, and
joins the second channel, which ends at Mycenae above the
Iroquois level. The lower waterway holds Round and Green
Lakes, which occupy depressions produced, as pointed out by
Dr. Gilbert, by the sinking of strata due to solution of the
underlying salt beds. The heads of these three channels are
higher than the High Bridge channel, and they probably
ceased to be effective earlier. Indeed they were probably the
first of the east and west series to be abandoned, excepting the
890 feet (“ Green’s”) channel, south of High Bridge.

Northward of the outlet series described above, lies another,
later, series of channels. The Syracuse sheet exhibits a broad
channel,. heading a mile east of Fairmount at 500 feet eleva-
tion, which passes eastward directly through the center of
Syracuse, where its elevation is only 400 feet. West of Syra-

HA. L. Fairchild—Glacial Lakes in Central New York. 263

euse they have not been mapped, but another and lower water-
course may be seen from the Auburn branch of the New York
Central railroad, heading in a cataract west of Marcellus station
and followed by the railroad eastward to beyond Camillus.
It is possible that one or more of the channels lower in ele-

_ vation than the Cedarvale channel may have established for a
short time an independent lake level, but such waters could
not have extended westward beyond the Ontario basin, as the
planes of these outlets are below that of Niagara.

-Upon the north sides of these channels the morainal drift is
heavy and the topography more irregular than shown by the
topographic sheets.

Lake Iroquois. |

These latest glacial waters of the Ontario basin have been so
long known that extended description is unnecessary here. The
_ name was given by J. W. Spencer, and Mr. Gilbert correlated
the shorelines with the col and ancient outlet at Rome. This
large and important body of water had a long life, during
which it matured its western shorelines and expanded to an
area larger than that of the present Ontario, as its surface was
far higher than Ontario, although the retreating ice barrier
during most of the life of Iroquois held possession of some of
the northern or northeastern part of the basin.

During the life of Iroquois the differential elevation toward
the north-northwest, or canting of the basin toward the south-
southwest, produced a shifting of the shorelines, a transgression
of the waters upon the southern shores and a recession from
the northern shores. This has been graphically shown by Dr.
Gilbert.*

The Iroquois shoreline was not matured east of Sodus
although abundant evidence of the water-action exists. South
and east of Sodus Bay the map gives only the general or
hypothetical position of the water line.

_ The lake became extinct by the opening of the St. Lawrence
Valley and lowering of the water below the Rome outlet. At
this time the altitude of the basin was such that the head of
the St. Lawrence channel, at the Thousand Islands, was below
sea-level, and a gulf of the ocean occupied the Ontario basin.
The differential uplift, however, soon barred the sea and
initiated the present lake and river.

*‘*The History of the Niagara River,” 6th Ann. Rept. Com, of State Res. of
Niagara, Albany, 1890.

964 TT. H. Means-—Soluble Mineral Matter in a Soil.

Art. XXVI.—A Rapid Method for the Determination of the
Amount of Soluble Mineral Matter in a Soil; by Tuos. H.
MEANS.

A RAPID and simple method for the determination of the
per cent of soluble matter which a soil contains has been found
very desirable where areas of soil are to be studied in any
detail. Particularly is this true in the arid and semi-arid
regions of the West where accumulations of soluble salts have
been the cause of injury to valuable tracts of land. The ordi-
nary methods of chemical analysis are entirely too laborious
for the examination of a large number.of samples, and, besides,.
chemical analyses cannot be made out in the field. In field
studies it is of great advantage to know the salt-content of the
soilathand. The accumulation of “alkali,” wherever irrigation
has been practiced a few years, has pointed out the great neces-
sity of studying the conditions of the soluble matter before
water isapplied. Enormous sums of money are being invested
in irrigation, and a great deal of this expended capital has
been, or will be, lost unless new systems of irrigation and
drainage are devised. ‘There can be no question that the study
of the soil, the alkali and the methods for the prevention of
its accumulation, must form an integral part of every irrigation
engineering problem. Not only must the surface of the soil
be examined, but the per cent of salt found in the subsoil to a
depth of 30 or 40 feet should be known. In order that under-
ground maps of the salt may be prepared with much greater
ease than is possible with ordinary means, the following
method is suggested.

The electrical conductivity of water varies with the amount
of dissociated salt that is present. The method is practically a
means of determining the specific electrical conductivity (or
resistance) of the solution between the soil grains. When this
is known, the amount of any salt which will produce this same
conductivity may be calculated from published tables of the
conductivity of salt solutions.*

In order to determine the effect of the solid grains of the
soil upon the resistance of the solution, the following experi-
ment was arranged. A hard rubber cell with parallel sides of
metal was partly filled with a solution of known specific
resistance. Then as much dry soil was added as would be
taken up by the solution, forming a saturated soil. The
resistance was in all cases increased. A. correction was made
for the soluble matter contained in the soil, and the increase in

* Kohlrausch, Wied. Ann., 1879, vol. vi, p. 145.

the results:

3 T. H. Means—Soluble Mineral Matter in a Soil.

265

resistance, due to the addition of the solid non-conducting par-
ticles, was therefore determined. In this way a large number
of samples were examined and it was found that the increase
- in resistance was nearly the same for soils of very different
textures. The following table contains the most important of

Soil and character.

ae yy es

oe ve (.

a

Ssonanooea lege To

‘ae

Early Truck soil, Md. Coarse sand -------.
)

Pe Risive sol, Blacksburg Va, ‘Heavy clay
6c “e |

Red Land, 5 6, Baal a be ott Neg A |
| “6 tc 6c 5

Sea Island, Cotton soil, 8.C. Fine sand _-_-_!

Per cent

water when resistance.

saturated.*

31°84
35°36
69°41
72°49
4 00
36°35
32-75
25-17
17-94

70 per cent of water for saturation.

|
|

| Initial

|
|
|
|

Ohms.

ee

Resistance

after soil |Factor of
texture.

is added.
Ohms.

1°81
is
1:90
1:85
1°83
1°88
i 8
1-83
1°78
Average

These soils, as seen in the table, vary from the light sandy
lands, which require less than 18 per cent (of their dry weight)
of water to: saturate them, to the heaviest of clays, requiring

Determinations were

made with two or three amounts of water in some of the soils
in order to strike as near as possible the point of saturation.
The last column is the ratio between the resistance after the
soil was saturated and the initial resistance of the solution.
the resistance (between parallel electrodes) of a saturated soil is
multiplied by this figure (the factor of texture), the resistance
of the solution between the soil grains will be found.

The salt determination is performed in practice as follows:
_ 1. Saturate the soil with distilled water.
ration can very easily be found and, with a little practice,
duplicate samples can be mixed to within a small error.
_ 2. Pack saturated soil in a cubic cell with two sides of metal
and take the resistance. Jtead the temperature, and correct

If

The point of satu-

resistance for temperature.t Multiply the corrected resistance

__ by the factor of texture, 0°55. This gives the resistance of the

soil solution alone in the cubic cell.

3. From the known weight of soil in the cell and the per-

centage of water required to saturate the soil, find the number
of cubic centimeters of soil moisture hotreed the soil grains.

In field work it is sufficiently accurate to take the average of

* Calculated on dry weight of soil.

+ Bulletin 7, Division of Soils, Department of Agriculture.

of Soils, Department of Agriculture, page 27.

18

Bulletin 8, Division

Am. Jour. Sci.--Fourto Suries, Vou. VII, No. 40.—AprRIL, 1899.

‘DDL
“519
‘527
"542
557
533
og
‘DAT
063
548

966 T. H. Means—NSoluble Mineral Matter wn a Soil.

several laboratory determinations of this quantity. The amount
of soil moisture being known, the cell factor (the figure by
which the resistance must be multiplied to obtain specific
resistance) for this quantity can be found.*

4. Multiply the resistance found in (2) by the cell factor and
obtain the specific resistance of the soil solution. For dilute
solutions the per cent of salt present is nearly inversely pro-
portional to the resistance. In a study of the alkali soils of

the Westt where the salts are of rather uniform chemical com-

position, it is of great advantage to have a more accurate
method of calculating the results. In such studies, a solution
of the average composition of the salts is made up and its
resistance-salt curve determined. In this way the per cent of
salt by weight can be reckoned for any resistance. Such a pro-
ceeding was followed in some studies of alkali soils in Montana.{

This method offers special advantage in the study of alkali
soils. The apparatus necessary can readily be carried in the
field. A convenient form of Wheatstone’s bridge adapted to

field use has been designed by Mr. Lyman J. Briggs of this

laboratory.

-The auger for sampling the soil is in general use in the study
of surface geology.§ The only other apparatus necessary are
the cubic cell, thermometer, mixing dish and bottle of dis-
tilled water.

This method offers a ready means for the detailed study of
irrigated districts. Two men can examine from sixty to one
hundred samples of soil in a day and in this way cover con-
siderable ground, obtaining sufficient data for the construction
of salt maps of the district.

U.S. Department of Agriculture, Division of Soils.

* Ostwald’s Physico-Chemical Measurements, translated by Walker, page 225.
Note.—If a cubical cell is used—such a cell 4° internal dimensions has given me
the best results—the cell factor will vary inversely as the number of cubic centi-
meters of solution present.

ee and Means, Bulletin 14, Division of Soils, Department of Agriculture.

Loe. cit.

§ Darton, Am. Geol, vol. vii, page 117-119.

| Whitney and Means, Bulletin 8, Division of Soils, Department of Agriculture,
page 11.

C. 8. Hastings—New Type of Telescope Objective. 267

Art. XX VII.— Ona New Type of Telescope Objective especially
adapted for Spectroscopic Use; by CHARLES 8S. HASTINGS.

THE ordinary achromatic doublet as invented by Dolland in
the last century, is, as is well known, very far from complying
with the condition implied in its name. For telescopes of
small aperture, or even for those of very considerabie aperture,
if a ratio of focal length to aperture as great as that customary
with Dolland be employed, the defect in color correction
is neither very conspicuous nor very harmful in ordinary
use. Bntif the apertures are very iarge, as in our modern
astronomical instruments, or if the length be reduced relatively
to the diameter of the objective, this defect of secondary color
aberration becomes very conspicuous and reduces the optical
efficiency of the instrument very materially. The maximum
inconvenience of the defect, however, falls upon the spectro-
scopist, who finds that, although the optical efficiency of his
instrument is independent of the wave-length of light which
he happens to be observing, the instrumental adjustments must
undergo frequent changes for adaptation to different portions
of the spectrum. Another familiar and obvious consequence
of the secondary color defect is the impracticability of adapt-
ing the same instrument to purposes of both eye and photo-
graphic observation.

It follows, therefore, that the solution of the problem, appar-
ently first seriously undertaken by Fraunhofer, namely, to
devise an absolutely color-free objective, is, and has long been,
of continuously increasmg moment. Fraunhofer failed; but
unless I greatly misapprehend the meaning of his own record
of his work, the effort led directly to the discovery of the
Fraunhofer lines and to the beginning of spectroscopy. It is
true that nowhere, as far as appears in his published writings,
- does he state that this was his aim; but in view of his very
extended experiments in varying the constitution of his glasses,
of his studies of the minute dispersion characteristics of various
substances, and of the extraordinary skill and conscientiousness
in perfecting an instrument which has possessed no other error
of importance since his unequaled contributions to the art of
telescope-making, few will question the validity of the
inference.

Doubtless many investigators since Fraunhofer’s time have
attacked the same problem, but, so far as I am aware, without
any recorded success until the writer showed, in a paper pub-
lished in this Journal, vol. xviii, p. 429, that there were certain
glasses, unfortunately not then procurable, which would yield

268 C. 8. Hastings—New Type of Telescope Objective.

in a triple combination an objective entirely free from color.
Since then the extraordinary increase in number of materials
at the command of the optician, resulting from the labors of
Dr. Schott of Jena, led the writer to return to the problem
with the results which were published in a paper in vol. xxxvil
of the same Journal, in 1889. A general method of dealing
with all such problems was there developed and a number of
triads were indicated which would yield the most favorable
results. It is true that those given included a phosphate glass
which was believed by the makers to be permanent and has
since proved not to be so; but it was distinctly stated that the
table exhibited a considerable number of other promising com-
binations, and the general method of recognizing them was
pointed ont. 3 .

It may properly be stated here that the latter paper also
contained a general discussion of interesting double combina-
tions, one of which promised to be of great value to spectro-
scopists ; but the inability of glass makers to supply large disks
of the materials in question proved an unforeseen difficulty.
Still, the writer employed this construction. for a number of
years for his spectrometer and only displaced it recently by an
improved type of objective. Professor Keeler has also
employed the same construction, made by Mr. Brashear with
my aid, satisfactorily in spectroscope work.

The experiments with triple combinations which followed
the paper last named met with an unforeseen difficulty. The
particular triad which promised most in theory proved to have
- in one of its members a perishable glass. This might possibly
have been used by covering the.objectionable material by a
more permanent glass cemented to it, but this course is not
without risk, and certain defects, to be noted later, are not so
readily eliminated if this method be chosen. The only practi-
cable course seemed to lie in replacing this material by one
beyond suspicion, and much time was spent in investigating
the possibilities of this meams. It was found, as appears from
the paper cited above, that there was no difficulty in selecting
triads which would meet the analytical condition, insuring
complete elimination of color and subject to practical limita-
tions as regards permanency ; but the necessarily greater curva-
tures of the lens surfaces introduced a new source of imper-
fection, namely, chromatic difference of spherical aberration.
Of course, this, like all other errors, is present to a greater or
less degree in all optical instruments which depend in any way
upon refraction for their action. In telescopes, however, this
error has never been sutliciently great to betray itself to the
users, although clearly indicated by theory. Gauss, indeed, a
long time ago showed how to reduce this particular error to a

4
"

~

is
—

SS es ee Ce ee
> ‘ a 2° pbanie oe gee

C. S. Hastings

New Type of Telescope Objective. 269

residual of a higher order of minuteness, but the fact that his
construction has never come into use isa most convincing proof
that the error is quite negligible as compared to other defects
inherent in the ordinary construction. But when we try to
make a color-free triple objective after the methods of the
paper of 1889, we find that the defect in question becomes of
great moment. Especially is this true if we prescribe the
cementing of the objective so that there shall be only two free
surfaces. Such an objective, if corrected as regards spherical
aberration for light-waves of mean length, would have strong
positive spherical aberration for the red and negative for the
violet ends of the spectrum. It is true that this defect might
not prove very obvious for a telescope which is to be used only
for objects which are approximately white, bnt it would be
intolerable in spectroscopic use. The obvious method of
reducing the error is to increase the ratio of the focal length to
the aperture. This method, however, would introduce such
serious structural and mechanical difficulties, and so far reduce
the convenience of handling all spectroscopes to which it might
be applied, that it seemed to me quite impracticable.

As a possible means of securing the end in view, convinced
as I am that its importance warrants any amount of labor, I
lately turned to a cousideration of the possibilities possessed by
a combination of four varieties of glass. The investigation
is necessarily somewhat laborious as appears from the unusual
conditions imposed from the outset; but the time expended in
attaining complete success was short compared to the protracted
investigations which led to a definitive rejection of the triplet
as quite inadequate. In short, I have constructed an_ objective,
consisting of a quadruple combination of silicate flint, borosili-
cate flint, silicate crown and barium crown, which possesses all
the properties demanded. It has but two free surfaces, the
four lenses being cemented together. With an aperture of
one-tenth the focal length, its focal plane is rigidly the same
for all wave-lengths, from that of the Fraunhofer line A to
that of K, while it is sensibly free from chromatic differences
of magnification and ot spherical aberration. With its perfect
color correction, the well-known (but ordinarily overlooked)
chromatic aberration of the eye becomes very sensible. This,
however, I have eliminated by means of a specially devised
ocular, so that in my. instrument there is no reason why its
length may not be made permanently invariable. One notable
advantage in the construction will appear at once to all spec-
troscopists: wave-surfaces from the collimator being rigidly
plane for all wave-lengths, the adjustment of the prisms for
minimum deviation—provided always that their faces are
accurately plane—ceases to be of importance. Thus a con-

270 0. 8. Hastings—New Type of Telescope Objective.

struction, which must have occurred to every one who has
seriously studied the theory of the spectroscope, in which
the last prism of a train is of half the angle of the remainder
and silvered on the back, so that the light retraces its course
through the train, becomes entirely practical. Indeed, my
experiments with the new telescope lead me to prefer a con-
struction of spectroscope in which the collimator and telescope
are set at constant angle, and the prisms, arranged as above,
are alone movable. This is the familiar construction of the
grating spectroscope.

Although the objective described above consists of four
lenses, I imagine that a cemented system of five lenses would
in some cases be preferable, especially in relatively large aper-
tures; but there is no doubt in my mind that four kinds of
glass are sufficient and, unless greater structural complexity is
admitted, necessary for the ends defined.

Should the construction meet my confident expectation and
supply the spectroscopist with an optical instrument combining
the merits of a reflector with the greater merits of a refractor,
it will be convenient to give it a characteristic name suggested
by its properties. These are, as given above, chromatic differ-
ences of focal distance, of focal length, and of spherical aberra-
tion, all reduced to practically zero, together with a minimum
possible number of free surfaces. As such an objective is the
same in its action upon light of all wave-lengths, I propose to
call it an esokumatic system.

Mr. Brashear of Allegheny, who made for me the prisms for
the study of these glasses, as well as scores of others, and who
has shown unfailing good nature and constant readiness to
lend me his efficient aid in all my optical investigations, merits
my unstinted acknowledgments. I have promised the neces-
sary calculations if he is called upon to carry out for others
this difficult piece of optical work which has yielded so
much satisfaction to the writer.

Yale University, March, 1899.

Pirsson—Phenocrysts of Intrusive Igneous Rocks. 271

Art. XXVIII.—On the Phenocrysts of Intrusive Igneous
Locks ;* by L. V. Pirsson.

THE igneous rocks, excluding those which have solidified as
uncrystallized glasses, may be roughly divided into two classes,
the granular, composed of mineral particles of approximately
similar size, and the porphyritic, consisting of minerals exhibit-
ing a more or less perfect outward crystal form which lie em-
bedded in a so-called groundmass made up of much smaller,
usually anhedral,t+ particles or even wholly or in part of glass.
A rock belonging to this latter class is a porphyry and the
larger embedded crystals are called “ porphyritic crystals” or
more conveniently “ phenocrysts,” a term proposed by Iddings
which has obtained general recognition among geologists.

It is with respect to the origin of these phenocrysts in a certain

' lass of igneous rocks that it is proposed to treat in this article.

Withont going into details, it will be sufficient to recall that
_ many petrographers have expressed the view that the pheno-
_ erysts of porphyritic rocks, by the very fact of their size, fre-
quent perfection of crystal form and contrast of character to
that of the groundmass in which they lie embedded, are older
than this groundmass, have been formed under different physi-
eal and therefore different geological conditions, in short, not
in the place where they now are, but at great depths, and hence
they are frequently spoken of as “ intratelluric” in origin.

The writer has been unable to find in the literature, that
any sharp distinction has been drawn between the origin and
character of the phenocrysts of ¢ntrusive igneous rocks and
those of the extrusive ones or lava flows. This has probably
arisen from the belief that phenocrysts are always formed, as
stated above, at much greater depths than the place in which
they are now found, and that therefore their origin is similar
in both cases. It would thus make no difference in regard to
their origin and character, whether the ascending magma
which contained them was thrust in intrusive masses into the
upper portion of the earth’s crust or poured out on the surface
in flows of lava.

In the present article, however, it is to be carefully noted
that the writer confines himself to the phenocrysts of the in-
trusive rocks, as it is intended to show that not all phenocrysts
are intratelluric in the sense that they have been formed at
much greater depths than they now occur in, but that on the

* Read before the Geological Society of America, December, 1898.

+ i.e., having crystal structure but not crystal form.

¢ Rosenbusch, Mass. Gest., 3d ed., 1896, p. 553; Michel-Lévy, Struct. et Class.

des Roches, 1889, p.1 et seq. See also discussion by Zirkel, who does not follow
this view, Lehrb. der Petrogr., 2d ed., 1893, vol. i, p. 737 et seq.

272 Pirsson—Phenocrysts of Intrusive Igneous Rocks.

contrary in many cases they have been formed in place, and
are of contemporaneous origin with the other constituents of
_ the rocks, and the nature of the proofs is partly such that they
can be applied only to intrusive rocks, and partly they have
not been observed in lavas. :

The idea that phenocrysts are not necessarily intratelluric
has been expressed by previous writers, and Zirkel* indeed
gives a very full discussion of the subject, while Crosst also
mentions this belief in alluding to orthoclase phenocrysts which
are found in certain laccoliths but not in the peripheral portions.
But the subject does not appear to have received in general
the attention that its importance merits, and the writer is
therefore led to present his own observations bearing on this
point, and the deductions he has drawn from them as a contri-
bution to the subject. They have been gathered during the
progress of the geological work in Montana, by Mr. W. H.
Weed and the writer, and the details have mainly appeared in
our publications on that region, but they are now gathered
together and generalized for use upon this question. Similar
proofs of course occur elsewhere, and they are indeed of
general application. |

Meaning of intratelluric.—The term “ intratelluric” needs
a moment’s attention. It is probable that different writers
have used it with somewhat different ideas, differing as to
degree, not as to kind. Nowhere has the writer been able to
find a precise definition by an author of what his idea of intra-
telluric is, but it is certain from the context, that in all cases,
as regards tume and niveau, it means an earlier period and
greater depth of the magma than that in which it came to
rest. Thus in a laccolith formed under a great. thickness of
sediments, the phenocrysts would be spoken of as the con-
stituents formed in the intratelluric period, that is a lower,
earlier period of the magma and in contrast to the groundmass,
which is never spoken of in these eases as intratellurie.

Whatever exact ideas writers may have had regarding this
term, it may be taken as well assumed that it means a deeper and
therefore earlier stage of the magma in the earth’s crust, than any
in which the intrusive porphyritic rocks have been formed.

Two classes of phenocrysts.—Phenocrysts may be divided
into two general classes. First, those which are found only
as distinct porphyritic crystals and do not occur as a mineral
component of the groundmass, like the micas and hornblendes
of many acid rocks, and the olivines of many basic ones; and
second, those minerals which are formed both as phenoerysts
and also in the groundmass, like the quartz and feldspar of

* Lehrbuch der Petrographie, vol. i, p. 737 et seq , 1893. |
+ Laccolitic Mountain Groups, U. 8. Geol. Surv., 14th Ann. Rep., 1895, p. 231.

y)

Pirsson—Phenocrysts of Intrusive Igneous Rocks. 273

- acid rocks and the pyroxenes of basic ones. The second class

| may be termed recurrent phenocrysts, meaning that the
| mineral composing them occurs both as a phenocryst and in

the groundmass; Rosenbusch suggests the term in speaking
of the recurrence of mineral formation. The other, or first
class will be called here monogenetic phenocrysts. This dis-
_ tinction is of importance, as it is chiefly with the recurrent
_ phenocrysts that we shall have to deal. |
_ The proofs that all phenocrysts of intrusive igneous rocks
are not intratelluric, but have in many cases been formed in
place, and contemporaneously with other constituents may be
_ divided into two classes, those which are megascopic, that is
_ to say, geological or may be observed in the field, and those
_ which are microscopic, or to be gathered by a study of thin
sections. The geological proofs will be considered first.

~Evidence of contact zones.—In the case of many laccoliths,
dikes, sheets and other intrusive masses of igneous rocks of
the more acid feldspathic kinds, it has been observed that

~ while the main rock masses are highly porphyritic, often contain-

ing phenocrysts of great size, they pass into peripheral zones,
which may attain considerable width, in which such pheno-
erysts are entirely wanting. The border zone is often observed
_ to be dense, aphanitic and totally without phenocrysts either
_ great or small—the constitnents are all of one period of forma-
_ tion and allotriomorphic in structure.*

But if the magma contained phenocrysts before coming to
rest, they should occur in the border zone as well as elsewhere.
It will not do to suggest two separate intrusions in these cases,
for the geological “mse en place,” and the whole character of
the occurrences, utterly forbids such a supposition. Nor will
it do to say that they previously existed in the border zone,
but were redissolved. For the border zone being the part
cooled first and most quickly, if they previously existed in it,
then there is especially the place where they should occur. If
any re-solution should take place, it would be in the center, not
at the sides. Nor will it do to partially compromise the
matter by saying that in the original magma they were very
small and grew greatly after the magma came to rest.

No doubt this often does happen, but in many cases, as
stated above, there are none in the contact zone, nothing to
represent the thickly swarming masses of them in the central
portion. In these cases the writer sees no escape from the
conclusion, that the phenocrysts were formed in place. The
evidence is most common and striking in regard to phenocrysts
of the recurrent type, but it has been also observed in those
of the monogenetic type. :

* Zirkel mentions similar instances, loc, cit., p. 740.

274 Pirsson—Phenocrysts of Intruswe Lgneous Locks.

Evidence of dikes and sheets.—In many eases we have clear
and indisputable evidence, that systems of dikes and sheets
have been formed by a single geological act; that is, filled
from a single parent magma, at the same time and under the
same conditions, and they have of course the same chemical
composition. An example occurring in the Judith Mts.* of
Montana has been recently described by the writer. The
attendant fringing sheets lying above many laccoliths may
also be cited as examples. Yet in many of these cases some
sheets or dikes may have large and well-developed pheno-
crysts, others none at all. The laccolith may be full of pheno-
crysts; its attendant sheets or some of them may be entirely
devoid of them. A sheet of tinguaite in the Judith Mts. con-
tains phenocrysts of feldspar as large as one’s land, but a nar-
row sheet of the same rock which the general geologic history
of the district shows was intruded at the same time, and which
has a similar composition, has only very minute ones arranged
in flow structure. In these cases, had they previously existed
in the parent magma, they should be found in both alike. ©
The conclusion in these cases is that the phenocrysts have —
formed in the places where they now are.

Evidence of fluidal phenomena.—tIn the ease of some dikes
and sheets it has been noticed that the phenocrysts of feldspar,
which have a flat tabular form, have a definite arrangement
throughout the mass, with the flat sides parallel to the bound-
ing walls, or in other words, as it is commonly called, a
“fluidal” or “flow structure.” Sometimes they are rather
irregularly distributed, collected in groups or in lines, and are
just as apt to occur close to the border zone, at the bounding
wall, as elsewhere in the dike. Such instances show clearly
the effects of tabular. plates being moved along in a liquid,
and they prove that the phenocrysts were formed in the
magma before it came to rest. How long before we are of
course unable to determine, but in these cases they may be
‘‘intratelluric,” so far as any definite meaning can be attached
to this term.

In other dikes and sheets, the phenocrysts have no oriented
arrangement in the mass, their orientation is wholly hap-
hazard, they occur evenly distributed throughout, or they may
be very small or of different character, or even wholly want-
ing in the border zones, and this in spite of the fact that they
may have a pronounced tabular development.

If in the former cases mentioned the evidence proves that the
phenocrysts were formed prior to injection, then the absence of
such evidence, that is to say its converse, proves just as clearly
that these flat phenocrysts were formed after injection when
the magma had come to rest or in other words in place.

* 18th Ann. Rep. U.S. Geol. Surv., pp. 551 and 572, 1898.

Pirsson—Phenocrysts of Intrusive Lgneous Rocks. 275

Moreover these flat phenocrysts, whether oriented or not, are
of great value in throwing light on another question in theo-
retic petrology—whether phenocrysts ever move in the magma
aiter it has come to rest; that is, whether they sink down in
obedience to gravity. In the cases mentioned it is clear that
they have not moved or they would have arranged themselves
in obedience to the laws governing the movement downward
of flat plates through fluids; they would be disposed with their
flat sides downward or perpendicular to the walls of the dike.
The writer has never seen such an occurrence nor been able to
find any description of one, and until such a case has been
found the evidence is against the idea that phenocrysts sink
through gravity and evidently because the magma is too viscous.
They could not assume such a position in an upward moving
magma because the motion must be differential and would arrange
them in the opposite direction ; hence this orientation would
show that they had moved and not the magma. The thicker
accumulation of phenocrysts at the bottom of an intruded
sheet, which has sometimes been noticed, would not prove they
had fallen through gravity unless they had this arrangement,
_ since that may merely show that the physical conditions neces-
sary for the production of phenocrysts were more favorable
and operative there than in the upper portion.

To return again to the formation of phenocrysts in place,—
from the evidences given above, when we find that in dikes
and sheets flat tabular phenocrysts such as the feldspars of
many acid rocks or the large mica tables of certain basic ones
are scattered without orientation, we may conclude that they
have been formed in place and are not “intratelluric.”

Luidence of porphyritic granites.—\t is well known that
many granites contain large porphyritic feldspars, often of
great size. The granites may have so coarse a grain that in
the absence of these feldspars no one would think of them
as other than typical granular rocks. The same is also true at
times of other rocks besides granites. The evidence that
these afford is chiefly of a negative kind, but it is useful to
consider in this connection. It is evidence not so much of the
tyme and place of the formation of phenocrysts as of the cause
and manner of formation. To treat these questions fully
would lead us into the discussion of the “ mse en place” and
the structure and classification of igneous rocks, which would
be beyond the limits of this article. It will therefore be suffi-
cient to ask one or two questions. Are these large phenocrysts
of granites to be considered intratelluric or not, or is granite
itself and all its constituents to be regarded as intratelluric ?
The difference between the porphyritic and the granular struc-
tures has been regarded by some, so far as their origin is con-

276 = Pirsson—Phenocrysts of Intrusive Igneous Rocks.

cerned, as a difference not alone of degree but of kend. If so,
what is this precise difference and why should granular rocks
ever possess a porphyritic structure ?

What the significance of these crystals really is according to
the author’s belief will be shown later on.

Microscopical evidence—We row come to those proofs
which are to be seen in the minute internal structures of the
phenocrysts when seen in thin section under the microscope.
They may be divided into two classes, external, or to be seen in
the minerals surrounding the phenocryst, and enternal, to be
seen in the minerals which the phenocryst encloses. We will
consider the latter first.

Internal evidence.—While the majority of phenocrysts are
homogeneous and free from inclusions, in a large number of
cases they are not, but contain the sther rock minerals. These
are usually those of an older period of formation, such as apa-
tites and ores, and their explanation appears simple, but in
some instances the phenocrysts may enclose al/ the other rock
minerals including crystals of the same mineral as they them-
selves are composed of. Thus large orthoclases are seen enclos-
ing other orthoclases and quartz. If the enclosed mineral has |
the same composition and a similar orientation as the pheno-
cryst, it becomes of course an integral portion of the erystal
and the enclosure cannot be detected, but it often does not and
is then readily identified. In these cases it is impossible to
avoid the conclusion that not only was the phenocryst formed
in place but contemporaneously with the other minerals. At
such times the included minerals are scattered throughout the
whole extent of the phenocrysts, but often they are found only
in outer zones, indicating that the inner portion had formed
first and that later the phenocryst was growing and expanding
while the groundmass was crystallizing and was thus including
it.* The ability of crystallizing material to orient itself over
considerable areas and enclose the other constituents is well
illustrated in the poikilitic structure as seen in the quartzes of
rhyolites, the orthoclases of shonkinites and the hornblendes
and micas of the very basic rocks.

Fixternal evidence—But growing and expanding pheno-
crysts do not always enclose the other minerals of the ground-
mass with which they come in contact. On the contrary, like
all crystallizing substances, they endeavor to exclude impurities
and more commonly they reject and push them along as they
grow. Usually the minerals which make up the groundmasses
of porphyritic rocks have an isodiametric development, they
have therefore no orientation as regards form, and it is hence
difficult to tell whether such particles lying alongside pheno-

* These Montana instances give additional force to those mentioned by Zirkel
in his discussion of the subject, loe. cit., p.. 747.

nie dies aia a of Intrusive Igneous Rocks. 277

_ erysts have been rejected and pushed or not. Iu certain cases
the microlites of the groundmass have however an unequal
_ development—they may be thin flat plates like the feldspars of
| some trachytes and andesites or slender. rods like the egirites
| of some phonolitic rocks. In such eases it is common to
| _ observe that the microlites have a length orientation parallel to
| ach face of the large crystal forming the phenocryst, and not
only this but the larger the phenoeryst i is, the more evident
_ this becomes; they have been more crowded and the ar range-
/ ment is more evident. This applies of course only to those
| microlites in the immediate neighborhood of the phenocryst.
| Inthe case of the large feldspar phenocrysts of some Montana
tinguaites the crystals are covered with a perfect felt of simi-
larly oriented egirite needles.* The arrangement is precisely
what would happen if a scattered group of matches lying ona
_ table should be swept to one side by a book and was caused
without doubt by a quite similar process.

Such arrangements have been largely either overlooked or
probably confounded with fluidal or flow structures. The lat-
ter, caused by movements or currents in the crystallizing
magia, is however easily told, since in that case all of the con-

_ stituents of the period are alike affected, both phenocrysts and
microlites being drawn out into waving lines or streams, while
- in ease the orientation is produced by the expansion of the
phenocrysts the microlites are arranged parallel to each of its
crystallographic faces, have no orientation in the ifterspaces
and the phenocrysts themselves have also no orientation. This
expansion structure—to coin a term for it—around phenocrysts
may of course occur in combination with fluidal structure and —
merge into it, but typical cases of it are easily told and should
be more generally noted in petrographic descriptions.

This evidence then goes to help the general proposition that
some phenocrysts are formed in the places where they now are,
for they were expanding and growing vigorously while the
groundmass was crystallizing and a magma whose groundmass
is crystallizing has already reached the place it is destined to
occupy.

Summation of evidence —The value of evidence is cumula-
tive and increases, not in an arithmetical, but in a geometrical
ratio. A single fact may indicate definitely a certain by pothesis
as its explanation, but it may also be an exception to a general
rule. When we can bring a second fact to support it, the
probability of the correctness of the hypothesis is oreatly
increased, and when a considerable number of such facts are
found, all pointing to the same explanation, we may reasonably
consider it as proved. Thus, when we find that a dike con-
tains flat tabular phenocrysts, which show no orientation, and are
wanting in the border zone of contact and contain all the other

* This Journal, IV, ii, 1896, p. 19).

ae? tc: — ' ,
a Tee (ea wa} So - Baws ¢
~ 2 aes ’ ions

OT de eel re, Sen ee |

278 Pirsson—Phenocrysts of Intrusive Igneous Rocks.

rock constituents, and when this dike is one of a set filled by
the same injection which may or may not have similar pheno-
crysts, we are thoroughly justified in assuming that its pheno-
crysts are formed in place and are not intratelluric.

Evidences of intratellurve origin.—W e should now consider
the evidences on which is based the view that phenocrysts —
have been formed at great depths under wholly different con-
ditions from the other rock constituents. So far as the writer
can discover, they are these,—the contrast in size and erystal
form of phenocryst to the constituents of the groundmass, the
fluidal arrangement the phenocrysts sometimes possess and the
fact that they are often corroded or changed in character, that
is resolved.

But the contrast of form and size, as we shall presently
show, may be explained in quite a different way; the fluidal
arrangement certainly shows that the phenocrysts were
formed before the magma came to rest, but it does not show
how long before or that it may not have been just previously ;
while the corroded and altered crystals show a change of con-
ditions, not necessarily a change of place, and although this
may be a logical inference it is still not the only one.

From the facts at our command, therefore, it appears most
reasonable to say that in some cases we know that phenocrysts
have been formed in place; in other cases we know that they
were formed prior to injection, but how long and how far we
do not know, and that in the great majority of cases we do not
know whether they were formed in place or not. It is the
author’s belief that the majority are formed in place for reasons
which it is now proper to give. “ie

Conditions governing the formation of rock structure.—So
far as the writer can understand, the French believe in two
distinct and different periods of consolidation for all intrusive
rocks both granular and porphyritic; many if not most of the
Germans believe in one for the granular and two for the por-
phyritic; like some petrographers, the writer sees no necessity
for more than one in both.

The conditions governing the consolidation and erystalliza-
tion of an igneous magma are complex, and many of them,
perhaps all, are as yet ill understood. The fall of temperature
is the most important, but the chemical composition of the
magma, the influence of mineralizing vapors and the pressure
are also of great importance. Another important considera-
tion which has been greatly overlooked is increasing viscosity,
which depends on chemical composition and temperature. It
would be impossible to discuss all these in full, and the writer
therefore contines himself to a brief statement of his own views.

The greatest determinant in the formation of rock structure
is the ratio of time in the fall of. temperature between the

Pirsson— Phenocrysts of Intrusive Igneous ocks. 279

point where the insolubility and crystallizing moment of a com-
pound begins to the increasing viscosity. Suppose a fluid rock
magma, without crystallizations in it, comes to rest and the
temperature begins to fall. At a certain point compounds, the

| ferro-magnesian ones let us say, commence to separate. The

fall has been so gradual that only a few centers of crystalliza-

| tion have been set up. The magma retains its fluidity for each

_ period and each mineral has a considerable time in which to
_ grow, orienting and assimilating its substance around it or, to
coin a phrase, it has a long crystallization interval. It is theo-
_ fetically possible that if the conditions are stable for a long
_ enough period each mineral in turn may exhaust itself in the
magma before the next begins. Usually there are overlaps.
Thus the process goes on until each has had its turn, the rock
has erystallized and the granular structure is formed.

Other things being equal, the granularity depends on the
length of the crystallization interval. Im the case of the very
siliceous magmas we must take into account another factor, the
_ viscosity, which depends greatly on the included water vapor.
_ If this is present in large amount the fluidity of the magma is
- enormously increased, and with it the radius of action of each
_ erystallizing center, that is, the crystal has a free chance to
_ grow, orient material and expand. On the other hand, if there
is little or no water vapor present or it escapes before erystalli-
zation begins, the viscosity is greatly increased, the radius of
action of each center is small, and, although the crystallization
interval may be sufficiently long, the crystals are restrained in
their growth, new centers are set up and the rock is fine-
grained. Hence we may have coarse granular and fine granu-
lar rocks without porphyritic structure.*

In the case of the basic rocks, the included water vapor has
no such function, their viscosity increases at a vastly slower
rate for each degree in fall of temperature, hence the radius of
action for equal crystallization intervals is much larger than in
the acid rocks and we therefore find them quite coarse granular
under the same conditions which have produced fine-grained
structures in the acid ones. Other things being equal, the
granularity depends on chemical composition, but there is also
a limit to this, for it also depends on the number of crystalliza-
tion centers which have been set up.

Porphyritic structure.—We can now consider the por-
phyritic structure. Suppose that an acid magma is injected
with included water vapor and that the fall of temperature is
comparatively rapid. The vapors escape rapidly and with the
cooling and their escape the viscosity augments in a greatly

*See also Iddings’ Crystallization of Igneous Rocks, Bulletin Phil. Soc.,
Washington, vol. xi, p. 105, 1889. ;

280 Pirsson—Phenocrysts of Intrusive Igneous Rocks.

increasing ratio. This means shorter and shorter erystalliza-
tion intervals for the various components as their periods com-
mence. Thus the earlier ferro-magnesian components may
have time to grow to considerable size while the increasing
viscosity forces the feldspars and later components to be of
very small size. Thus we may have a porphyritic structure —
with phenocrysts of the monogenetic type as a result of a
single process of crystallization. It may be considered a sort
of arrested or retarded granular structure.

But phenocrysts of the recurrent type evidently demand a
further explanation and this is to be sought in two things.
First the writer believes that the influence of mass action is of
great importance and. that compounds which are present in
relatively large or predominant amount tend to set up centers
of crystallization before what might otherwise be the proper
period for that component. The rest of the process might go
on as noted above anda porphyritic structure would be formed
with phenocrysts of the recurrent type. Phenocrysts of the
recurrent type would then consist of those minerals which are
present in largest amount, and this is generally true. Such phen-
ocrysts might form at any time and the earliest would be crys-
tallizing during the whole period and might include the other
constituents or exclude and orient them as previously noted.

Second, the writer is inclined to believe that:too great regu-
larity in the period of commencing crystallization of the dif-.
ferent components has been ascribed to crystallizing magmas. |
That the different minerals ¢end to have their proper periods
is undoubtedly so, but they do not always have them, and, for
example, centers of feldspar erystallization may be set up at
different times resulting in porphyritic structure as mentioned
above. If we dissolve salt in water and allow the solution to
crystallize in different beakers, we may obtain under apparently
similar conditions a few large crystals, many small ones or a
mixture of large and small. The conditions are of course not
exactly similar but the differences may be very slight.

The more coarsely granular the rocks are, the more stable,
and even the conditions under which they have crystallized
and the less likelihood there is for the formation of pheno-
crysts—yet these do occasionally occur, as noted above in the
porphyritic granites. In the fine-grained rocks these conditions
are changed, the causes mentioned above become operative and
a porphyritic structure is liable to result, as it commonly does.
_ Thus it seems most reasonable to regard the majority of
phenocrysts as formed in place though in some eases they have
evidently formed previously while the magma was ascending.
Even in these cases they have probably been formed not long
before.

Yale University, New Haven, Conn , December, i898.

J. H. Pratt—Occurrence of Chromite, ete. 281

Art. XXIX.—On the Occurrence, Origin and Chemical Com-
position of Chromite; by J. H. PRart.

Introduction.—In a recent paper* on the origin of the corun-
dum associated with the peridotites in North Carolina,
attention was called to the constant presence of the mineral
chromite in these rocks. During the past summer the occur-
rence of chromite has been carefully studied and the occur-
rences in North Carolina have been examined in the field. As
a result of these examinations the author has been led to adopt
the theory that the chromite should be regarded as having been
formed at the same time with the peridotite, i. e., as having
been held in solution by the molten mass of the peridotite and
crystallized out among the first minerals as the mass began to
cool.

This theory is essentially the same as that advanced by
the authort+ for the origin of the corundum associated with the
peridotite rocks, and a similar line of reasoning has been used
to substantiate the theory proposed.

Investigations concerning the igneous origin of some of the
ores have been materially aided, during the recent years, by
the able experiments of Morozewicz{ and Lagorio$ and. by
researches that show us more clearly why we should regard a
fused mass of rock as a liquid, having similar properties to an
ordinary solution. The fused mass of rock is capable of hold-
ing different minerals in solution and as the molten mass begins
to cool, these minerals would separate out not according to
their fusibility but according to their solubility in the fused
mass. The more basic minerals being the more insoluble
would be the first to separate out, and as was mentioned in the
paper already referred to, this crystallizing or solidifying out
from the molten mass would take place first on its outer
boundaries, for here it would cool first. Convection currents
would tend to bring new supplies of material to the outer zone
where crystallization takes place...

Occurrence.—W ith the exception of alluvial deposits, chro-
mite has only been found in the peridotite and allied igneous
basic magnesian rocks, or in the serpentines which have resulted
from the alternation of these rocks. In the North Carolina
peridotites chromite occurs more commonly as scattered grains
or crystals, but it is also to be found inthe form of imbedded
masses near the boundaries of the lenticular bodies of dunite.

* This Journal, vol. vi, p. 49, July, 1898.

+ This Journal, vol. vi, p. 50, July, 1898.

t Zeitschr. fiir Kryst, vol. xxiv, p. 281, 1895.
§ Ibid., p. 285.

Am. Jour. Sc1.—Fourtu Series, Vou. VII, No. 40.—ApriL, 1899.
19

282 J. H. Pratt— Occurrence, Origin and

It does not occur in well defined veins, but is often in masses
or pockets which apparently have no relation whatever to
each other.’ But few writers on the occurrence of chromite
have described the relation of the chromite deposits to the
rock in which it is found. One or two have mentioned the
chromite as being found near the eastern boundary of the ser-
pentine or at the northern border of the serpentine belt, but
no definite description of the occurrence could be obtained.
The large deposits of chromite in North Carolina occur in the
peridotite rock and near the contact of this rock with the
enclosing gneiss. Also where there is but a small amount of
chromite either in small pockets or in grains or erystals, these
are more abundant near the contact and diminish in number
toward the center of the mass of the peridotite.

Where these large deposits of chromite occur there has been
no corundum or but very little found, and where we find the
large deposits of the corundum there is a scarcity of chromite.

This constant occurrence of the chromite in rounded masses
of varying proportions near the contact of the peridotite with
the gneiss, its occurrence in the fresh as well as in the altered
peridotite, indicate that the chromite has been held in solution
in the molten mass of the peridotite when it was intruded
into the country rock and that it separated out among the first
minerals as the mass began to cool.

As has been said, the peridotite (dunite) magma holding in
solution the chemical elements of the different minerals would
be like a saturated liquid, and as it began to cool the minerals
would separate or crystallize out, not according to their fusibil-
ity but according to the degree of their solubility in the molten
magma. The more basic minerals, according to the general law
of cooling and crystallizing magmas, being the less soluble,
would therefore be the first to separate out. These would be
the oxides containing no silica and in the present case would
be the chromite, spinel and corundum. ‘These minerals would
solidify or crystallize out where the molten magma first began
to cool, which would be at. the contact of the mass with the
country rock; convection currents would tend to bring new
supphes of material to the outer boundary which would deposit
its chromic oxide as chromites. This would account for all the
irregularities of the chromite deposits : their pocket-like nature,
the shooting off of apophyses from the main masses of the
chromite into the peridotite, the widening and pinching of the
chromite lodes, and the apparently non- relation or connection
of one pocket of chromite with another. There has not been
enough work done in the North Carolina chromite mines to
demonstrate the exact position and relation of the chromite
deposits to the gneiss or the country rock, and in the descrip-

oe Sa
a

Chemical Composition of Chromite. 283

tion of other chromite mines but little light has been thrown

on this point.

The chromite would be concentrated near the borders of the
peridotite in rounded masses, with offshoots penetrating into
the peridotite. The line of contact near the gneiss would be
sharp and nearly regular, while with the peridodite the con-

tact would be very irregular.

In prospecting for either chromite or corundum the largest

and richest deposits may be expected near the contact of “the

peridotite with the gneiss or other country rock.

been analyzed from
the following ieealiges: in North Carolina: Price’s Creek, six
miles southwest of Burnesville, Yancey County ; Webster,
Jackson County ; and Corundum Hill, Macon County.

Pure material for analysis, showing no impurities when
examined with the microscope, was readily obtained by hand
picking.

The mineral was fused several times with bisulphate of

potash, then taken up with hydrochloric acid and silica tested

for. Iron, aluminium and chromium were precipitated with
ammonia, ‘the precipitation being made at least three times.
Magnesium, calcium and manganese were determined in the
filtrates by the usual methods.

The precipitate of the mixed oxides was dissolved in hydro-
chloric or nitric acidand the excess of acid evaporated. Sodium
hydroxide was then added in excess and chlorine passed into the

hot solution. The solution was acidified and the iron and

aluminium precipitated twice with ammonia and weighed as
mixed oxides. These mixed oxides containing a trace of
chromium were fused with acid potassium sulphate, digested
with water and acidified. One precipitation was made with
ammonia to partially remove sulphates. This precipitate was
dissolved in hydrochlorie acid and treated as before, the iron
and aluminium being obtained free from chromium. The iron
was determined volumetrically.

To the filtrates containing the chromium, alcohol and hydro-
chloric acid were added and the solution digested for some
time. The chromium was precipitated as “hydroxide and
weighed as Cr,O,. To ensure the purity of the precipitate, it
was fused with four parts of sodium carbonate to one part of
potassium nitrate, the fusion taken up with water and tested
for magnesium.

A number of experiments were made to determine the ratio
of ferric to ferrous oxide, but they were all unsatisfactory. By
digesting the very finely powdered mineral in a mixture of hot
concentrated hydrofluoric and sulphuric acid, in an atmosphere
of carbon dioxide for half an hour, enough of the mineral was
mo posed to show the presence of ferric oxide.

284 J. H. Pratt—Occurrence, Origin and

The analyses were made by Dr. Chas. Baskerville* and Dr. —

H. W. Foote.t+

The result of the analyses were as follows:

Price Creek. Ratio. Corundum Hill.t Ratio. Webster§ Ratio.

CxO. 122. 959:20 386 57°80 377. 39°95 | -96 1m
AOE cae Tals 070 7°82 076 29-28 287m
FeO .... 25°02 347 25°68 ‘356 18°90 -192mm
MgO -.. 4749 sala 5°22 ‘131 eee
BIOL. 8320 pee 2°80 Bobi: ee

MnO_.... 92 SOEs eG GE bis ae

In the above analyses the ratio of the bivalent oxides to the

trivalent oxides is uniformly high and can probably be accounted

for by reason that some of the iron calculated as ferrous oxide
was in the ferric state, as was proved in the Webster chromite.

Taking enough of the Cr,O, and MgO to unite with the FeO
and AJ,O, respectively to form the molecules FeO. Cr,O,, and
MgO. Al on there remains approximately enough of the Or, O,
to unite with the excess of the MgO to form the molecule
MoO .Cr,0,. The nearer the ratio of the bivalent oxides
equals that of the trivalent oxides, the nearer the excess of the
Cr,O, and MgO equal each other. The inability to determine
the ratio of the ferrous to the ferric oxide in the above analyses
prevents the obtaining of a sharp ratio in the excess of the
Or,O, and MgO.

With the exception of two cases, in all the terrestrial chro-
mite analyses examined, alumina and magnesium were inyari-

ably present, varying from a small per cent in some samples to —

others that showed the presence of more than 20 per cent of
these oxides. In the above analyses and in most of the others
examined, it was noticed that the magnesia usually varied with
the alumina, those rich in alumina being correspondingly rich
in magnesia.

This constant occurrence of magnesia and alumina in the chro- —

mite would seem to indicate that the molecule of the mineral
now called. chromite is not pure FeO. Cr,O, but is a combina-
tion of the three isomorphous molecules, FeO. Cr,O,; MgO.
Cr,O, and MgO. A],O,.

But two analyses of chromite (terrestrial) have been found ~
that do not show the presence of magnesia and alumina, the ~

first a magnetic chrome sand from Chester, Pa., analyzed by

a eee

T. H. Garrett, in which all the iron is calculated as ferric —

* Of the Chemical Laboratory, N. C. Geological Survey.
+ Of the Sheffield Laboratory, Yale University.

t Baskerville. S Foote.

|| This Journal, vol. xiv, p. 47, 1852.

Chemical Composition of Chromite. 285

oxide, and the second chromite from Wache Island, West
‘Indies, analyzed by J. Clonet.*

I (Garrett). II (Clouet).

Beet 8 41°55 51°53
_ 4 eee abi 48°46
LOG, Se eee 62°02 puch
Mp pieeen de. ols 1°25

104'82 fase

_ A greater part of the iron in Garrett’s analyses is undoubt-

edly in the form of ferrous oxide, as is indicated by the high
per cent obtained. From the above, it is seen that a pure
chromite having the composition FeO. Cr,O, is not common
in nature.

The MgO. A],O, occurs nearly pure in nature as normal
spinel. Thenormal MgO. Cr,O, has not been found in nature,
but, as indicated from the above analyses, there is good reason
to believe that this molecule does exist, and we may expect to
find normal MgO. Cr,0, occurring asa definite mineral. Under
this theory of the composition of the chromite, the formule
for the three chromites described in this ‘paper would be as
ae

. Price Creek 10FeO . Cr,O,; MgO. Cr,O,; 2MgO. Al,O
3 Corundum Hill 9FeO. Cr,O,; : Cr,0O,; 2MgO. Al Oe
3. Webster FeO. Cr ‘0.; ) MgO. Cr,0,; - 2MgO. Al "OL

The first two formule will represent + ae those of
the majority of the chromites that have been found, and they
are approaching the normal chromite FeO . Cr,O, as their limit.

As the FeO .Cr,O, molecule decreases and ‘the MgO. Al,O,
increases the mineral would approach normal spinel MgO.
AJ,O, as its limit, the mineral picotite or chrome spinel being
representative of a mineral near the spinel end.

- M. E. Wadsworth,t in comparing the chromite with the
picotite associated with peridotites, says:

“Jt is probable that picotite and chromite belong to the
same mineral series, the term picotite bemg more commonly
applied to the freshest states, and that of chromite to those
forms more altered, and to the local aggregations arising from
the migration of the chromic oxide during the alteration of
the associated peridotite rocks. As a further extreme in
the alteration a change to a more or less magnetite occurs.’
These conclusions are deduced from a microscopical study of
these minerals.”

* Ann. Chim. Phys., vol. xvi, pp. 90-100, 1869; and Wadsworth’s Lithological

Studies, appendix, p. iv.
+ Lithological Studies, p. 184.

286 J. H. Pratt—Occurrence of Chromite, ete.

The author does not agree with Dr. Wadsworth that the
chromite represents an altered product of a mineral, of which —
picotite is a purer form. The chromite is a mineral which
suffers alteration but slightly and, as it is found at the present
time, represents the original mineral and not an altered form.
The difference in the microscopical properties can readily be
accounted for by the difference in the chemical composition.
With an increase in the ratio of the molecule, MgO. Al,O,
and a corresponding decrease in the molecule, FeO . Cr,O,, the
more translucent the mineral will become.

These two minerals belong to the same group and are closely
allied to each other, and they represent two primary minerals
and not different stages in the alteration of another mineral.

In the analysis of the Webster chromite, the largest per-
centage of MgO was obtained and in the calculation of the
ratios the formula was shown to be FeO.Cr,O,; MgO. Cr,0, ;
2M¢O.Al,O,, this being the highest ratio of the molecule
MgO. Or,O, of any chromite examined. The theoretical com-
position is here given, together with the analysis of the Webster
mineral :

Theory for
Webster chromite found. 2Mg0. Al.03;
MgO 5 Cr20; ; FeO. Cr.0s3.
Seow es 2 §80'95 40°90
PAURO Coa Rak: 29°28 30°44
He whl ek, 13°90 10°75
Mie Qs 2 ais. 17°31 Ee Sh|
100-44 : 100-00

This analysis is similar to that described by Bock* for a
“magnochromite” from Grochau, Silesia, which contained
Cr,O, 40°78, Al,O, 29°92, FeO 15°30 and MgO 14-00.

In appearance this Webster chromite is different from any
that has come under the author’s observation, being much more.
of a coarse-grained appearance than ordinary chromite.

In order to designate this Webster chromite and others of
similar composition, the author proposes the name, Mvtchellite,
in honor of the late Professor Elisha Mitchell of North Carolina.

N. C. Geological Survey, December, 1898.

* 7s G. Gedeanye pis 04 Sites

J. T. Norton, Jr.—Titrations by Sodium Thiosulphate. 287

Art. XXX.—The Influence of Hydrochloric Acid in Titra-
toons by Sodium Thiosulphate, with special reference to the
Estimation of Selenious Acid; by JOHN T. Norton, JR.

[Contributions from the Kent Chemical Laboratory of Yale University—LXXX.]

_In the method of Norris and Fay* for the iodometric deter-
mination of selenious acid, advantage is taken of a direct and
uuique action of sodium thiosulphate upon selenium dioxide
in the presence of hydrochloric acid. Most excellent results
are claimed for this method; but the explicit statement of the
originators of the method, that the amount of hydrochloric
acid present does not influence the result, provided the titra-
tion is made at the temperature of melting ice, is so extraordi-
nary in view of generally accepted ideas in regard to the
interaction of hydrochloric acid and sodium thiosulphate, as to
suggest the necessity of careful investigation of this point.

Pickering,t in his investigation of the reaction between iodine
and sodium thiosulphate, has shown that more iodine is required
to oxidize the thiosulphate as the proportion of hydrochloric
acid increases. He ascribed this effect to the formation of. a
sulphate, apparently, by the increased activity of the iodine,
but the more rational explanation is that, although some sul-
phate is ultimately formed, the thiosulphate is first partially
decomposed into free sulphur and sulphur dioxide. Finkenert
and Mohr§ also mention the decomposing effect of free acid
upon sodium thiosulphate.

The sodium thiosulphate used in the following experiments
was taken in nearly decinormal solution and was standardized
by running it into an approximately decinormal solution of
iodine, the value of which had been determined by comparison
with decinormal arsenious acid made from carefully resub-
limed arsenious oxide. In the experiments of Table I the
solutions were stirred continuously and kept at a temperature
of from 0° to 5° C., while the thiosulphate ran into the acidified
liquid. The volume of the solution, though fixed at the be-
ginning as given in the table, was considerably increased during
the operation by the melting of the ice. Titrations were con-
ducted as rapidly as possible to avoid the separation of sulphur,
which is likely to occur, especially when the acid and thiosul-
phate are present in large quantities. A perusal of the table

* Am. Chem. Jour., vol. xviii, pp. 703.
+ Jour. Chem. Soc, vol. xxxvii, pp. 135.
¢{ Anal. Chem., 6 Aufl, pp. 620.

§ Titrirmethode, 6 Anfl., pp. 279.

288 J. T. Norton, Jr.—Ilnfluence of Hydrochloric Acid

shows that the influence of the hydrochloric acid upon the
thiosulphate depends chiefly upon the amount of the thio-
sulphate present and afterwards upon the degree of dilu-

TABLE I.
Volume of N a2S20s Volume of ay iodine used in titration.
liquid at approximately 10
beginning of ~ faa a ———--~- —
titration. 10 °... HCl=nene: =lem*?. =5 cm? =—10 em®.
em’. em’. em?. em’. em?, em*,
100 30 30°25 ) 30°75 30°76 31°2
200 fy) 30°22+,| mean== 30°21 30°56 314
300 ‘2 30°20 \ 30°22 30°22 31°03 30°9
400 a 30°21 | 30°20 30°20 30°55
500 ef 30°20 J 30°20 30°21 30°55
100 25.2 2 25°32 25°98 25°70
‘200 a 25°28 | mean== 25°34 25°40 25°45
300 = 25°29 p25: 27 25°41 25°38 25°83
400 < 227 25°24 25°30 25°63
500 A 25°22 | 25°23 25°40 25°30
100 20 20°15 20°17 20°33 20°23
200 oe 20°20 | mean= 20°13 20°27 20°23
300 re 20°21 220% 20°15 20°20" aur y
400 4 20°20 20°10 20°27 20°07
500 20°10 | 20°10 20°17 20°13

tion and its own absolute quantity. Thus when 30™ of
sodium thiosulphate were employed the effect of 10™ of acid
is marked at all dilutions within the range of the experiments ;
the effect of 5°™* of acid is inappreciable only at a dilution of
from 400 to 500°, and when 1™ of acid is employed the
effect is only perceptible at a volume of 100. When 25°
of the thiosulphate were used the influence of the acid is less
marked; for at a dilution of 500° the effect of 10°™° of acid
is not seen, and 20° of the thiosulphate may be present at
any dilution down to 100° in the presence of as much as 10°™
of the acid, and even considerably more, as experiments not
included in the table indicated. ,

The shght discrepancies which appear occasionally in the
table were due, no doubt, to unavoidable differences in the
time of action.

This influence of time upon the reaction between sodium
thiosulphate iodine and hydrochloric acid comes out clearly in
the following series of experiments, in which the thiosulphate
was run into the acidified water, cooled to a temperature of
from 0 to 5° ©. by means of ice, the solution being allowed to

‘

in Titrations by Sodium Thiosulphate. 989

stand 5,10, and 15 minutes. Sulphur was thrown down in
nearly every case. . |

TABLE II.
Volume of the Na.S.03; psi the eee
q liquid at HCl approximately Volume of 10 iodine used
beginning of sp er.(112) NV in titration after standing.
titration. present. 10 é 5 min. 10 min. 15 min.
em?, em?. em?, em?. em?, cm3,
200 10 30 30°80 31°30 HS pigs 7
< : 4g 25 25°50 26°00 26°30
ZO o 20 20°30 20°70 20°68

The results of the table emphasize sufficiently the necessity of
proceeding as rapidly as possible with the titration of sodium
_thiosulphate by iodine in presence of hydrochloric acid, when
_ the thiosulphate is present in considerable amount. As would
_ be expected, the effect of temperature upon the reaction is also
marked. In the following experiments the sodium thiosul-
phate was run into the acidified water, previously heated to
the temperature indicated, and then titrated with iodine.

_ From these results it is plain that the conditions under which
considerable amounts of sodium thiosulphate are titrated in
presence of hydrochloric acid must be carefully guarded when
accuracy is a consideration. It is also apparent that in all
_ eases the temperature should be reduced as nearly to 0° C. as
possible and rapidity: of titration by the iodine is an essential.
_ So long as the thiosulphate present does not exceed 20° of

the 2 solution, rapid titration in cold solution proceeds with

_ fair regularity in presence of hydrochloric acid up to 10° of
_ the acid of sp. gr. 112. When, however, the amount of thio-

TABLE III.
1 = Volume of a
Volume of Na.S.0; a)
liquid at HCl. approximately. jodine used in ti-
_ beginning sp. gr. (112) Temp. N Bees trations at differ-_
_ of titration. taken. Centigrade. 10 "ent temperatures.
em?*. cm?, C. em?. em*.
400 10 6° 25 23°52
““ ““ 99 ° «“ 23°73
‘x &s 34° ce 24°35
“ & 49° -- 24°5
ce “ce 54° ce 95
és “ 64° = 26°1

4

290 Jf. T. Norton, Jr.—Influence of Hydrochloric Acid

sulphate is greater than 20°" of the = solution, care as to the

restriction of the acid and dilution of the solution becomes a
necessity. Fortunately, in most analytical processes involving
the use of the thiosulphate it is possible to add that reagent
from the burette to the solution to be acted upon, so that it
is destroyed normally as fast as it is introduced and the danger
of interaction with the acid does not occur. In the process of
Norris and Fay, however, the method involves the addition
of an excess of the thiosulphate to the solution of selenious and
hydrochloric acids, and thus the conditions prevail which
demand care as to the relation of the acid, the thiosulphate
and the degree of dilution. I have experimented, therefore,
with this process under varying conditions.

The process of Norris and Fay for the iodometric determi-
nation of selenious acid consists briefly in treating the solution
of that acid in ice water, in the presence of hydrochloric acid,

with an excess of a a solution of sodium thiosulphate and

titrating back the excess of the thiosulphate with iodine.
Four molecules of sodium thiosulphate act, apparently, upon
one molecule of selenious acid according to a reaction which
the authors propose to study.

The selenium dioxide used was made by dissolving pre-
sumably pure selenium in nitric acid and evaporating to dryness.
The residue was then treated with water, and a little barium
hydroxide was added to remove any sulphate which might be
present. The solution was then filtered and the filtrate
evaporated to dryness. The residue was mixed with four or five
times its volume of dried pulverized pyrolusite, and the whole
was put into a porcelain crucible and heated. The sublimate of
selenium dioxide was carefully collected on a dry watch-glass —
and put into a drying bottle as quickly as possible. The pyro-
lusite prevents any reduction of the selenium dioxide to
selenium and the product consisted of beautiful long white
needles. This method of preparing the selenium dioxide,
which has been used for some time in this laboratory, avoids
contamination of the selenium dioxide by nitric acid or water,
resulting from the decomposition of the latter, which would
be possible in case this reagent were employed in the final
sublimation, as is recommended by Norris and Fay. The
hydrochloric acid used was of asp. gr. 112, as recommended
by the originators of the process. For the experiments of Table.
IV the dilution at the beginning was fixed at 400°, and this
was increased in every case by the melting of the ice used to
cool the liquid. A glance at the preceding part of this paper

in Titrations by Sodium Thiosulphate. 291

_ will show that at this degree of dilution the hydrochloric acid
_ present has the least effect.

~ TABLE IV.
Volume
Amount HCl at begin- Excess
SeO. sp.gr. oingof Na.S.0; SeO.
taken. 112 ee titration. employed: found. Error.
grm. em?. cm’. cm?. orm. orm.

1) °0616 10 400 2°28 0°0625 0°0009 +
2) "0628 Ha @ he Ss 0'0631 0°0003 +
3) *0508 ~~ 3 11°4 0°0511 0°00038 +
4) 0587 Fe “a 12°8 0°0594 0:0007+ mean
5) "0807 4 of 15°3 0°0813 0°0006 + :00005 +
6) -0633 ‘i - 20°85 0°0638 0°0005 +
7) °0682 25 i Aa | 0°0685 0°0003 +
8) -0779 a s 1.55 0:0788 00-0009 +
9) °0465 = * 18°93 0°0469 0°0004+

ws ae ee o : ———
LN OSU

These results, while not so good as those obtained by Norris and
Fay, are satisfactory and show that at this degree of dilution the
process is accurate. These results accord closely with those
contained in Table I. Ata dilution of 400°™° or in the pres-
ence of only 20° of sodium thiosulphate in excess the
hydrochloric acid present had no perceptible effect. Of course,
it must be kept in mind that the hydrochloric acid acts only
upon the excess of thiosulphate which is not taken up by the
selenium dioxide. The slight constant plus error in these
results cannot be accounted for by errors in the standards ;
they were all carefully determined. Another preparation of
selenium dioxide was made, starting with pure selenium c¢are-
fully precipitated by sulphurous acid, before putting it through
the course of treatment previously described, and the results
obtained by the action of the sodium thiosulphate recorded in
Table V agree closely with those of the preceding table.

TABLE V.

Amount HCl H.O Na.S.0;3

SeO. Sp. gr. at in SeO.
taken. 112. beginning. excess. found. Error.
grm. em’. em’, em’, grm. grm.
(1) 0562 10 400 9°52 "0566 "0004 +
(2) ‘0651 25 4 11°20 "0655 0004+

The next step was to determine the effect of diminishing the
dilution and of varying the strength of acid. The following
table gives the results of my experiments.

EE

292 JS. T. Norton, Jr.—Injfluence of Hydrochloric Acid

TABLE VI.
Volume
Amount of H,O0 HCl Excess
of SeO, atbegin- sp.gr. of Naz SeO2
taken. ning. (112). S20s. taken. Error.
erm. em? em? em? erm. erm.
(1) °1042 200 9) 24°16 "1041 ‘0001 —
(2)s2061 14 yas? 10 13°3 ‘0611 ‘0000+
(3) 0850 “ia 10 21°9 "0828 "0022 —
(4) :0757 Pe 25 13:07 "0749 “6008—
(5) "0540 2h 21°02 "0522 ‘0018 —
(6) "0674 300 3) 10°04 ‘0679 "0005 +
(7) 2416 400. 5d Loco "2424 °0008 +

It is apparent that at the dilution of 200° we run into difficul-
ties, and the greater the excess of thiosulphate present the
greater is the error. When the amount of sodium thiosulphate
exceeds 20°" a reduction in the amount of acid to 5°™* is plainly
of advantage, as is shown in a comparison of Exps. (1), (8),
and (5), and is not disadvantageous at larger dilutions and
with smaller amounts of the thiosulphate, as shown in Exps.
(6) and (7). The necessity of placing some limits on the
method of Norris and Fay has now, I think, been estab-
lished. The excess of the thiosulphate must be carefully regn-
lated, as well as the temperature. If one has knowledge of the
approximate amount of selenious acid in solution, this is not a
matter of great difficulty, and things should be so arranged

that no more than 20°™* of the — thiosulphate should ever

be present in excess of the amount necessary to reduce the
selenious acid. If this limit —amounting to 0:0400™ of SeO,—
is placed upon the thiosulphate, so much as 10™ of hydro-
chloric acid (sp. gr. 112) may be present without endangering
the accuracy of the process, provided the solution is diluted to
400°" at the outset; if only 5°™* of hydrochloric acid are
present, the volume at the beginning may be reduced with
safety to 200. At all events, 5°" of the hydrochloric acid
are amply sufficient to bring about the reaction between the
thiosulphate and the selenium at any dilution within the range
of my experiments. With these precautions taken, the process
of Norris and Fay is simple, rapid and accurate ; without them,
as the experimental results indicate, errors of considerable
amount may enter.

According to the method of * Muthmann and Shafer, the
determination of selenious acid is effected by the simple addi-
tion of potassium iodide to the acidulated solution of selenious

* Berichte d. d. Chem. Gesell., xxvi, 1008.

Vee aR eee Rss ees eae

———
ant

eae me

in Titrations by Sodium Thiosulphate. 293

acid, and the iodine set free is titrated with sodium thiosul-
phate. In this procedure the thiosulphate is taken up by the
iodine as it is added to the solution, so that the danger of any
action between the thiosulphate and the acid is out of the
question. It was shown in a former paper from this labora-
tory * that this simple method is inaccurate on account of the

incompleteness of reduction in the cold and in presence of the

iodine evolved. In a later article also from this laboratory +
it was shown that selenium may be completely precipitated
and determined with accuracy gravimetrically provided the
amount of potassium iodide employed is enormously in excess
of that theoreticaliy required. This suggests naturally the trial

_ of very large excesses of potassium iodide in the process of

Mnuthmann and Shafer. The details of experiments made in
this manner are given in the following table. |

TABLE VII.

HCl

SeO. Vol. of Sp. gr. SeO.

used. Kel solution. (112). found. Error.

erm. erm. cm? em? erm, orm.
(1) °0553 10 150 10 0558 0005 +
{2) -0574 3) r vs "0567 ‘0007 —
(3) 0683 5 ? * ‘0683 ‘0000+
(4) "0487 5) if * "0484 "00038 —
(5) ‘2617 10 “ = "2589 "0028 —

Jt is obvious that for small quantities of selenium dioxide the
accuracy of the process is very much increased by the use of
large amounts of iodide, though, of course, the difficulty in
reading the end reaction due to the presence of precipitated
red selenium still remains; but the process is still inaccurate
when large amounts of selenium dioxide are employed.

In conclusion I wish to thank Prof. F. A. Gooch for his
kind advice and assistance.

* This Journal, Gooch and Reynolds, vol. 1, 254.
+ Peirce, this Journal, vol. i, 1896, p. 416.

294 TLurner—Rock-forming Biotites and Amphiboles.

Art. XX XI.—Some Rock-forming Biotites and Amphiboles ;*
by H. W. TurNneER, with analyses by W. F. HILLEBRAND,
H. N. Sroxszs, and WILLIAM VALENTINE. |

THE minerals referred to in this paper were separated from
the rocks containing them by means of the Thoulet solution.
The analyses were made in the laboratory of the U.S. Geo-
logical Survey, the object being to furnish a basis for calculat-
ing the molecular composition of the rocks containing them.
All of the material was fresh and sensibly free from foreign
particles. The powder analyzed in each case was examined
microscopically and such impurities as were found are given.
Samples of minerals as well as chips of the rocks containing
them will be sent to any one who is investigating rock-forming
minerals, upon request. The specimens came from the Sierra
Nevada in all cases.

Biotite.

Biotite No. 2136 is from the biotite-granite at the base of E]
Capitan in Yosemite Valley. It is quite black in color, as seen
with a hand lens, but under the microscope is brown. This
biotite-granite forms large areas in the granitic complex of the
Sierra Nevada, and is quite uniform in appearance and compo-
sition at nearly all points where I have seen it. The biotite
which it everywhere contains as an essential constituent, is
quite the same in color and degree of pléochroism in the many
thin sections examined, and the analysis given below may be
fairly assumed to represent in general the composition of the
biotite of this granite. Biotite-granite 2136 contains quartz >
plagioclase >orthoclase> biotite > magnetite > titanite >apatite
>zircon, the last four minerals being merely accessories. The
aplitie granites that cut the other granites in dikes likewise
contain a very similar biotite in small amount.

A microscopic examination of the material analyzed, as well as
the mere fact that phosphorus pentoxide is present, showed that
there were apatite prisms mixed with the biotite. The analysis
was therefore recalculated and the P,O, (-20) taken out together
with the CaO (-26) required by the P,O, to form apatite. The
H,O below 110° C. is likewise left out in this recalculation
since it was no doubt adventitious.

Biotite No. 1751 S.N. is from a quartz-monzonite collected
about 1 klm. southwest of Blood’s Station, Alpine County, in
the Big Trees quadrangle. The material analyzed was found,
on microscopic examination, to contain occasional white grains,

* Published by permission of the Director of the U. 8. Geological Survey.

‘OD GOL J¥ pourmiejed »

a puriqel[iH| puesaqor[iy | euuereA | saxo1g eulueyeA | ouTyTE]e A. Sune e A |S "atone ee ysk[euy

€8-66 16-66 00-001 00-001 06-66
S 10. 70: | Ul. (Ne aon eee tae ae es ‘T= we8hxo
aS 06-66 96.66 |. 00-001 09 001 11-001 OL =e eo [B10 7,
"~» _—_—_—_ —_—_———s SSS ss SS | eee —_-__ooOoOoOoOoOoO eo een Oe eee
‘eS ee eae Or Sass. he gouen aoedl "OK 98: oe ae aa i ere ae a
e. See ee as 2 epee 7 eer = GEE aod Sie eee em malia ge estas a ean? Be eee raps
eee we a potas alg ES ES eal Gn Ue tcc ene ONY gob rage GuGtis. 7S eee a sea ars OM ae Peet Sa eee Tels

G0 ee es ee wae ee euou aes euou Ol. UR WR? ae Bt Gs pi OE Se ee SRE Se he ee aes |
Se pyar te ee ee 92-£ Sy Aes 9€-F 0g. 19.2 VaCgs Mehes sey ae ‘0 .O1T eAoge OH
Sx80-1 eS: Esa] OG ee Seve OU OU S iihe akt as a Sika er os aay es "0 .O1T MO[8q OF H
w% ze a. pooien Ui} 90B 1) bY 90BI4 90B 1} DOC tee ok ay ee oe ees oy Ory
XS Bete ee cp | 6a-8 te F9-€ ge Gt oe fell tage tees ae et “-" OFeN
S Cdn eee ae a 08-8 eL-€ (ay) 80-F GE-6 QRGR ler she ae at ae
ae Oe Wn ee eal SOG Se toe) Sain at OL-GT bg. Z8 6 GIGS ee ye mes Site Os
RQ ae pias eee ie Sai 9%. 70: aoed} coats @ le set Soe ee org
Dd» Cl ae. ee ey ee (3) lo | Oe eae ae eoeB Ty GO: 9uou euou AR aes SI A ce ee alg OAS
S ite A Beh ee So 8h 19-€ a6. 09-2 69. 9 cpa ap ep ee eee oro
= OO Sig) Pe Ey ett geo eee ca Wi ese a Se aise (oat BF oa 009 pur OIN
S Cig: Cade peepee Oe Slr bl. G1. 08. GL: SF eee oe ae ree = OU
ie G0 rl ele Pon) OGLE ues : 19-€1 Te. GL FL TE aaa ae gromaier seks Se Oem
ne GO) Fie a or ER OST = el ae : 69-4 69. 09-4 Tce. |S pe eae - ie ce "O%og
S G0. MO ore pala eee eel a os ee ep ibe ee Rae it oes cols meer ae etal ee eee py So a SOFA
R eb anre te aay Sa fae ge tee | gree ee plea Sie eae TE De = Seen aa edtias Si 5.5. dltea @h ces i oe oe ae ee eke SO°19

| Ck Ss. oe (ernie: Sh Saas £6-GT 06-ST 18-81 COC TS aoe lea we Maer ee SO°lV

Pa euou ee day OO ie ee, ee eg a et ee CU ee ae wee I LG ee ery ®OIZ
: OT Gee tel Sas ae GUIdse el saaae 19-2 aa ETT Gilg etheg oe pee eee SOML
S G1.G8 G.19 09.98 16:99 29 9¢ 80-1) 20-98 VOSGES rt sa “Veer eee S018
: ‘S'g B98 “CPLL AS ‘98-1% oyuRIS] “poyeNojeoar

oor | SPAT SION | oarnougy | OHMOZTOM | “TALL OMOIKT) "Snore ania 9g18 OMOLg

-24.18N()

7 = =

296 Lurner—Rock-forming Biotites and Amphiboles.

probably feldspar. The impurities, however, form such a
minute portion of the material that the analysis may be con-
sidered as sufficiently correct for use in calculating the con-
stituents of the granitic rock of which it forms a part. The
mica is quite black in.color, as seen with a hand lens. Under
the microscope it is a darker brown than the biotite of the
biotite-granite. This biotite is characteristic of the granodio-
rite and quartz-diorite series, and of the quartz-monzonite. It
contains more magnesia than any other of the biotites analyzed
but nearly the same as amphibole 2652, also from quartz-mon-
zonite. Dr. Hillebrand examined specimen 1751 and the |
biotite from it for vanadium. He found that the rock con-
tained -012 per cent of V,O, and the biotite -066 per cent.
Quartz-monzonite 1751 is composed of plagioclase >quartz>
orthoclase>biotite>amphibole. There are present as acces-
sories titanite, apatite, and iron ore.

Biotite No. 1748 S.N. is from a pyroxenic¢ gneiss from the
south bank of the North Mokelumne River, about 1 klm. up
stream from the mouth of Bear River, in the Big Trees quad-
rangle. The material analyzed was found, on microscopic
examination, to contain occasional. grains of amphibole, but
these probably do not sensibly affect the result. This biotite is
reddish-brown in color and quite similar to the biotite so often
found in contact metamorphic rocks. Dr. Hillebrand subse-
quently tested the material for vanadinm and found gneiss
1748 contained -08 per cent V,O,, and the biotite from it -127
per cent. Gneiss 1743 is composed of plagioclase, reddish-
brown biotite, greenish-brown amphibole, quartz, pyroxene,
titanite, magnetite, apatite, and pyrrhotite, the relative abund-
ance being approximately indicated by the order in which the
constituents are enumerated.

Biotite No. 2652 is from a quartz-monzonite from a block by
the Tioga road southeast of Mt. Hoffmann in the Mt. Dana
quadrangle. The amphibole 2652 analyzed is from the same
block, and both were separated as well as analyzed by Dr.
Hillebrand. The block above referred to was far richer in
amphibole and biotite than is usually the case with the quartz-
monzonite of the region. There is thusa possibility that these
minerals may differ somewhat in composition from the same
minerals where they occur evenly distributed through the
magma. An examination of the powder analyzed shows on
one slide a very few foreign particles, namely, apatite and
amphibole. The apatite should of course be deducted from the
analysis as well as the water below 105° C. Dr. Hillebrand
found that the mica contains no rare earth, zine, or metals of
the H,S group. The specific gravity of biotite 2652 at 21° C.
is 3°05. Quartz-monzonite 2652 was not analyzed, but for

Turner—fock-forming Biotites and Amphiboles. 297

_ purposes of comparison No. 2179, given in the table of
amphibole analyses, may be substituted, as this is from the
same area and represents the average rock.

A comparison of the biotite analyses given in the table with
those of the rocks from which the micas were obtained, does
not suggest any very definite relations. These micas are all
_ quite similar in composition. It may be noted, however, that
the biotite richest in alumina and alkali is from the most acid
rock of the series, the biotite-granite (No. 2136). This rock
contains no amphibole. Biotites 1751 and 2652 are richer in
magnesia. both of these are very similar in composition as
well as in color and pleochroism, and both of them are asso-
ciated with amphibole.

Analyses of Amphibole and the containing rocks.

| | | Quartz- Amphibole |
Amphibole Gabbro Amphibole | monzonite | Donegal | Donegal
ore iy RatO. a 2 ape 2199. granite. granite.
{eee 46°08 47-27 47-49 66°83 47°25 58°04
eres a 92 1°21 *B4 eee EE e508
ah Bae) 2p rs ee ae eee none 04 =U, ees are 24:
_ eae 10°56 20782)" ",| 707 15°24 5°65 16°08
ee Fk PaOb waee™ |. os es, epee See Fees
a Se "02 04 i) ae fo Be 2) eS
ee Srey 2°81 1°85 4°88 2°13 19-d1 8°27
ee 8°30 4°26 10°69 1°66 "94 0°45
|e 15 trace 51 10 Ley Oo» a ina
Ot a none ih 1 SLRS ye IPs i at ee
See 12°64 13°02 1-92 3°59 eee deg 6°52
. none trace none ‘03 Ag ie ean toy es
a none none none ay! Biase eae o
aa 1440 | 6°44 13°06 1°63 11°26 2°94
_. ee 34 “22 ‘49 4°46 1:04 2°21
ea 1°62 2 75 ‘75 3 10 98 4°65
a none none trace eaemrny es: Ste ete St
1,0 below 110° C.__ 17 08 + TCIETS iglal s ames Se Bs Bale 1 Sea
[,0 above 110° C.__) 197 1:27 1°86 oad an ae Be | aie AS
a 18 14 none mises yh o's & Pe, = clo
a re none sh Fike trace Bc eeea = ate
ee ee ee a SR a mime ‘20 ee sos, acre a ee
eee mone |?) 2th Ae
Ee ee Lrace) sie eRe fe a (i ee eae pes pe
. none 25 Eat ‘06 * 5 720 REDS EE ee 2 ogee
Sa 99°99 99°86 100705 10082 | 99°69 100°28
cies cm wa Valentine Stokes | Hillebrand| Valentine Haughton | Haughton,

No rare earths or zine present.

Am, Jour Sor1.—Fourrsa Series, Vou. VII, No. 40.—AprIL, 1899.

An examination of the powder of amphibole 1970 shows
that it contains a few grains of labradorite but otherwise appears

* Cr.0; and V.0; determined by Dr. Hillebrand.
+ Dried at 105° C. before analyzing.

298 TLurner—Lock-forming Biotites and Amphiboles.

to be quite pure. The phosphorus pentoxide (P,O,) shown
in the analysis must, however, be ascribed to apatite, and this
should be deducted from the amphibole analysis together with
a corresponding amount of lime. In addition minute black —
grains may be noted in the amphibole itself. These are prob-
ably magnetite, but the amount thus included must be small,
since the powder was treated with an electro-magnet and the
more highly magnetic grains removed. No titanite or other
titaniferous mineral were found in the thin sections of gabbro
1970 or in the amphibole powder. The titanic oxide (TiO,)
shown in the analysis therefore probably belongs to the amphi-
bole. The pleochroism is strong in brown and greenish brown
colors with large extinction angle.

Amphibole-gabbro 1970 is from Beaver Creek about 18 klm.
east of Big Trees P.O. in the Big Trees quadrangle. It is
composed chiefly of labradorite and amphibole but contains a
little pyrite and pyrrhotite.

Amphibole 2652 is exceptionally pure. It is pleochroic in
dark green tints with large extinction angle. There are some

minute black grains included in the amphibole. These are ~

probably magnetite. One foreign particle only was noted ina
slide of the powder. As before stated, this amphibole comes
from the same block of quartz-monzonite as biotite 2652.
The specific gravity of amphibole 2652 at 21°5° C. is 3-203. 7
Analysis 2179 is that of an average quartz-monzonite of the ~
region. It is well known that the metasilicates of rocks rich —
in alkalies are sometimes themselves rich in alkali. This relation ~
does not appear to obtain in the case of amphiboles 1970 and ~
2652, the former of which comes from a basic rock poor in
alkali, and the latter from an acid rock rather rich in alkali,
yet there is more alkali in the amphibole from the basic than
in that from the acid rock.

There is also added to the table an analysis of an amphibole
from the Donegal granite* as well as one of the granite itself. —
It is evident from the analysis of this granite that the rock is 7
really a quartz-diorite or granodiorite and not a true granite. —
The biotite from the Donegal granite is remarkable for the
very high content of ferric jron—19°11 per cent.

* Haughton, Quart. Jour. Geol. Soc. Lond., vol. xviii, 1862, p. 416.

*

O. P. Hay—Species of Saurocephalus. 299

Arr. XX XII.—On one litile known and one hitherto unknown
species of Saurocephalus; by O. P. Hay.

THE fish Saurocephalus lanciformis was first described and
named by Dr. Richard Harlan in 1824.* This description and
the accompanying figures were reprinted in 1835 in the same
author’s Medical and Physical Researches.+ The specimen on
which the genus and species were based had been collected
about twenty years previously, by Lewis and Clark, at some
locality probably in northeastern Nebraska. It consisted of
the greater portion of the left maxilla; but was described by
Harlan as belonging to the lower jaw. He also regarded it as
having belonged to a reptile allied to Lehthyosaurus. Louis
Agassiz first recognized the ichthyic nature of the remainst
(although he confounded them with an entirely distinct spe-
cies); and his conclusions were confirmed by Richard Owen.$
Dr. Leidy| corrected Agassiz’s errors, and gave more accurate
descriptions and figures of the maxillary than had been fur-
nished by Harlan.

No remains of Harlan’s species, other than the maxillary
referred to, have hitherto been described. Dr. E. W. Hilgard§
has reported the species as occurring in the Vicksburg group
of the Eocene, but the identification was undoubtedly erro-
neous. Dr. William Spillman** has also included this species
in his list of fossils belonging to the Tombigbee greensand of
the Cretaceous at Columbus, Miss. Although this identifica-
tion is less improbable than the former, we have nothing to
confirm its correctness.

Notwithstanding the scantiness of the material belonging to
the type species, our knowledge of the genus Saurocephalus
has been greatly increased through the descriptions of closely
related and more perfectly preserved species. For this addi-
tional knowledge we are indebted to Cope and Newton, and
more recently to Alban Stewart, of the University of Kansas.

For sometime I have had in my possession some remains
which on examination prove, in my judgment, to belong to
Harlan’s species. This material was collected for me in the
region of Butte Creek cafion, south of Wallace, Kan.; and
the horizon is undoubtedly that of the Niobrara Cretaceous,
My material consists of both the mandibles, the right maxilla,
the pterygo-palatine arch and a few other bones.

The maxillary (fig. 1) is rather short and deep. The portion
belonging in front of the palatine condyle is missing ; but the
condyle itself is present. The alveolar border is somewhat

* Jour. Acad. Nat. Sci. Phila. [1], iii, pp. 331-337, pl. xii, figs. 1-5.

+ Med. Phys. Res., pp. 362-366, pl., figs. 1-5.

Poiss. Foss., v, p. 102. § Odontography, p. 130, pl. 55.
Trans. Amer. Philos. Soc., 1857, xi, pp. 91-95, pl. vi, figs. 8-11.
“| Report Geol. and Agric. Mississippi, 1860, p. 142. *® Op. cit., p. 389.

300 O. P. Hay—Species of Saurocephalus.

curved, and is occupied by compressed sharp-edged teeth. Of
these there are present twenty-eight; but if we restore the
bone, as we can safely do, I believe, by aid of Stewart’s figures

Figs x2
2 Sis.

et a

of S. dentatus,* we may conclude that there were originally
thirty-four teeth, possibly one or two less. The root of the
most anterior tooth has been exposed by the fracture, and its
fang is seen to be distinctly faceted ; so that it presents just
such an appearance as the tooth of S. lanciformés figured by
Leidy.t The roots of teeth situated more posteriorly whose
fangs have been exposed by a tool are similarly faceted. Cope
statest that S. lanczformis is to be distinguished from his S.
arapahovius, by the lack of facets on the roots of the teeth
of the latter. |

Leidy estimated that the maxilla in his hands had supported
only twenty-six or twenty-eight teeth, and he was probably
correct. That maxilla, a larger one than the one in my posses-
sion, seems to have been broken just behind the palatine con-
dyle. If now we take from Leidy’s drawing the width of the
bone at this point and apply it to the alveolar border, we find
that it includes ten teeth; the width of my own specimen
includes thirteen teeth. Itis not impossible, however, that the
specimen figured by Leidy had been broken away some little
distance behind the condyle. At any rate, I do not believe
that the difference of a few teeth, other things being alike,
would justify us in regarding the specimens as belonging to
different species.

As in the case of the original specimen there is a shallow
groove running along the mesial surface of the maxilla, about
5™= from the alveolar border, and from this groove foramina,
one for each tooth, enter the bone.

Depth of maxillary at palatine condyie -_--- 33am
Distance from anterior end of palatine condyle
to hinder end of maxillary). 2-2: oe 85™™

The right mandible is shown in fig. 2, five-eighths the natu-
ral size and showing the mesial surface. The alveolar border
is straight and supports thirty-four teeth, of which those occu-
pying the middle of the border are the largest. In general,
they are larger than the teeth of the upper jaw. ‘The line

* Kan. Univ. Quart., vii, p. 25, pl. i, figs. 3a, 4a.
+ Trans. Amer. Philos, Soc., xi, pl. vi, fig. 9. {Cretaceous Vertebrata, p. 216.

O. P. Hay—Species of Saurocephalus. 301

which spans thirteen teeth in the maxilla spans ten in the

dentary. At the proximal end of the mandible there must

have been a process of the dermarticulare, as in related forms ;

but in the specimen figured it is hidden by the overlying
Fig. 2. x3

eae

eeratohyal, which is not shown in the figure. At the anterior
end of the mesial face of the dentary there is found a broad
surface, rough with processes and pits, an indication that the
two dentaries were strongly bound together. The extreme
anterior end of each dentary is occupied by a surface to which
was evidently attached such a predentary as Stewart has de-
scribed as belonging to several related species. A groove and a

row of foramina are present on the median face of the dentary.

Length of alveolar border ._-_..-- ee
Length of mandible from cotylus._ 130
Depth of mandible at last tooth... 56

Depth of mandible at symphysis -- 34

Fig. 3 represents, five-eighths the natural size, the pterygo-
palatine arch seen from within. A triangular piece is missing
from the anterior end, and the lower end of the ectopterygoid,

Fig. 4. x4

pq, is defective. As I interpret the bones, the arch is remark-

able for the large size of the palatine, pa. While the sutures
which are represented in the figure are very distinet, I am
wholly unable to find one separating the entopterygoid, ey,
from the metopterygoid mt. pg. On the upper border of the

302 O. P. Hay—Species of Saurocephalus.

arch, at the point indicated by the line s, there appears to be an
indication of a suture. If such it is, it probably extends down-
wards to a point near the hinder end of the palatine. The
arrangement of the bones is quite different from that found
by myself in Xzphactinus.* |

At the lower border of the anterior end of the palatine
_ there is a broad surface, v, which was probably in contact with
an articulating surface on the vomer. The notch seen in the
anterior end is occupied by another articulatory surface, ma,
for the anterior palatine condyle of the maxilla. The anterior
end of the upper border furnished an articulation, pfe, with
the prefrontal, but this is elongated and rough, not broad and
smooth, as it is in Xzphactinus.

Anteriorly the palatine is thick and strong. On its outer
surface. this portion is finely vermiculated above, while the
lower portion furnishes a concave articulation for the condyle
of the maxilla. The general appearance of this portion may
be seen from fig. 4, which represents the palatine of the next
species. Below the concave surface for the palatine condyle
of the maxilla there is seen a broad rough surface which must
have been applied to the inner face of the maxilla. The
greater portion of this is wanting in the specimen shown in
fio. 4. Its limits are indicated by the dotted line. On the
outer face of the metapterygoid, from the highest point seen
in fig. 8 there runs downward and backward a sharp ridge
which evidently bounded the orbit below. The portion of the
metapterygoid above and mesiad of this ridge formed the floor
of the orbit. This indicates that the orbit was placed well
backward. I find no satisfactory evidences of the presence of
teeth on the pterygoid and palatine bones. If we shall add to
the maxillary the probable antero-posterior extent of the pre-
maxillary, we shall find that it is approximately equal to the
length ,of the lower jaw. Hence the latter did not project
beyond the upper jaw as it did in the case of those species
which Stewart has referred to the genus Sawrodon.

Two characters seem to distinguish Sawrodon from Sauro-
cephalus, viz.: the presence of notches, instead of foramina,
for the successional teeth and the projection of the lower jaw
beyond the snout of the fish. JI have been inclined to believe
that the presence of these two characters is sufficient to dis-
tinguish Sawrodon as distinct. However, I observe in some
specimens of this supposed genus that some of the notches
become closed into foramina; and we can easily imagine all
gradations between notches and foramina high above the alve-
olar margin. Moreover, it is probable that the other character
will fail. Recently Mr. Stewartt has published figures, with-
out description, of remains which he refers to Cope’s Sauro-

* Zoolog. Bull., ii, 1898, p. 39, fig. 7.
+ Kan. Univ. Quart., vii, pl. xvi, figs. 4, 5.

O. P. Hay—Species of Saurocephalus. 303

don phlebotomus. Mandible and maxilla are shown. Measure-
ments show that the maxilla, without the premaxillary, is
nearly as long as the alveolar border of the mandible, so that
it is almost certain that in this species there was no projection
of the dentary beyond the snout. It seems probable therefore
that Sawrodon must be abandoned.

I present here (fig. 5) the right maxilla and the premaxillary
_ (fig. 4) of another species of Saurocephatus, which I regard as
_ yet undescribed. It is especially distinguished from described
species by its elongated maxillary bone. To illustrate this, I
compare it with Mr. Stewart’s S. dentatus, which is itself a
species with a rather long maxilla. In & dentatus the total

per cent of the shorter maxilla. My species, therefore, prob-
ably had a relatively slender head and a larger mouth than had
S. dentatus.

In the maxilla figured I count alveoli for thirty-seven teeth ;
but in the maxilla of the other side, somewhat broken, the
teeth extend backward somewhat farther; so that there must
have been forty. At some time in the career of its owner the
right maxilla has been fractured obliquely across its middle,
and this accident has affected the neighboring teeth. One of
these has thus become exposed nearly half-way to the tip of
the fang. This exposure reveals the fact that the fang is
faceted, as it is in S. Janciformis. The great length of the
maxilla distinguishes this species from both S. lanciformis and
S. dentatus, and the facets on the teeth distinguish it from
Cope’s 8. arapahovius. Mr. Stewart has not described the
condition of the fang of the teeth of his 8. dentatus.

In fig. 5 p. ¢ represents the palatine condyle; p. c’, the
anterior palatine condyle which was applied to a surface like.
that shown in figure 3 at mz.

I propose to call the fish above described Sawrocephalus
pamphaqus.*

* Inde ruunt alii rapida velocius aura,

Pamphagus et oar
Ovid, Met., Bk. iii, 1. 209.

804 O. P. Hay—Species of Saurocephalus.

It has been supposed that the foramina, situated one opposite
each tooth and on the mesial face of the maxilla and of the
dentary, are for the transmission of nerves and vessels to the
teeth. Richard Owen* seems not to have so regarded these
foramina. He believed that they “lead to the cavities contain-
ing the germs of the successional teeth.” The latter probably
began their development in, or at the bottom of, these foramina ;
but they soon passed more deeply into the bone. In fig. 1 at ¢
there is found a developing tooth whose tip is on a level with
the row of foramina: but its root extends. high up into the
bone. Nerves and vessels entering the tooth by way-of the
foramina alluded to would have to take a very tortuous course.
The functional tooth immediately below the young tooth fig-
ured seems already to have suffered some reduction of its fang. —

The germs of the teeth of the Saurocephalide did not gain
a lodgment in the bones of the jaws in the same way that the
teeth of the higher vertebrates did. In the latter the fangs
were first planted in grooves in the dental borders of the bones ;
and we must suppose that these grooves, at first shallow, have,
in successive generations, deepened and become portioned off
to form sockets. In the Sawrocepnalide the teeth, developing
originally on the dental border, have gradually migrated away
from this border, on the mesial face of the supporting bones,
and, by means of the foramina described above, have made —
their way through the mesial wall of the sockets. The notches —
found in the species referred to Saurodon show the earliest
stages of this migration. :

The distinguished paleeo-ichthyologist, Mr. A. S. Woodward,
has recently kindly called my attention to a suggestion made
by Prof. E. D. Cope that the Saurocephalide are closely
related to the Chirocentride, represented by the large Chiro-
centrus dorab of the Chinese and Indian seas. I have unfortu-
nately had no opportunity to study a skeleton of this fish; but,
judging from the figures of the fish found in Cuvier and
Valenciennes, pl. 565, and in Day’s Fishes of India, pl. elxvi,
fig. 38, its external appearance must be much like that of the
extinct Xiphactinus. Nevertheless, we have no intimations
that the teeth of Chirocentrus are fixed to the jaws in any way
different from those of ordinary fishes. The fixation of the
teeth in sockets is an unusual thing among fishes; and this
character alone, it appears to me, is suflicient to remove
Aiphactinus and its allies from the Chirocentride, although
not necessarily to a great distance. I suspect that the Sauwro-
cephalide will, when they are better known, show distinctive
characters in the vertebral column also.

* Odontography, p. 131.

G. R. Wieland— American Fossil Cycads. 305

Art. XXXIII.—A Study of Some American Fossil Cycads.
Part Il. Zhe Leaf Structure of Cycadeoidea ; by G. R.
WIELAND. (With Plate VIL) .

Introduction.

It is generally conceded that the Cycads culminated in the
Jurassic, and formed the principal plant life of that time.
Nevertheless, the abundant and widely distributed fossil
remains of this vegetation consist almost entirely of discon-
nected and scattered trunks, leaves, and occasional fruits, which
have been described under separate generic and specific names.
So uniformly is this true that Count Solms-Laubach’* mentions
an individual of Zamites gigas, Morr., from the upper Jurassic
sandstone of Yorkshire, England, as the only eycad known to
him, whose stem with the leaves attached may be identified
with certainty. This specimen, now in the Paris Museum, is
figured by Saporta,’ and in leaf and habit is said to recall
Stangeria most. nearly. ,

There has, therefore, been no sufficient means of correlating
the various genera and species founded on isolated parts,—a fact
which now makes especially important a determination of both
trunk and leaf characters when united in a single individual.
For this reason it is fortunate that not only the leaf and its pre-
foliation, but also the accompanying floral characters, may
be completely determined in a number of the unusually per-
fect cycadean trunks now in the Yale collection, though but

a single one is considered in this paper.

Prefoliation of Cycadeoidea.

The subject of the present, as well as of the preceding, pre-
liminary study is afforded by the type specimen of Cycadeoidea
ingens, Ward, a photograph of which has been already given.
As previously mentioned, this fine male trunk bears on its
summit, above the series of flower buds, a crown of young
leaves imbedded in fine scales, or ramentum. It would seem
that, having formed its blossoms, this eycad was again prepar-
ing to put forth its energies in the unfolding of an additional
series of leaves, about twenty in number. Happening to be
fossilized at such a critical time in its life, under the most
favorable conditions, there has been preserved a wealth of
characters scarcely equalled by any Mesozoic plant of which
there isa record. This specimen alone affords all the material
required for a restoration.

* The numbers refer to the list of papers given in Part I of these studies.
This Journal, March, 1899, pp. 219-226, Plates IIT, IV.

306 G. R. Wieland—American Fossil Cycads,

A careful examination of the summit of the trunk showed that,
as found in the field, it had been subjected to the erosive action
of sand-laden winds which had partially exposed several trans-
verse sections of the tips of young leaves. It was found that
these belonged to a circularly placed series just ready to emerge
at a distance of from 5 to 8 emfromtheapex. Several pairs of
these leaves enveloped in ramentum were removed for study,
and the disposition of their parts with reference to the trunk
determined from longitudinal, transverse, and tangential
sections. :

As so removed, each rachis with its attached pinnules formed
an erect subcylindrical, or spindle-shaped, body, 6 em in
length, with a long diameter of 15 mm coinciding with the
radius of the trunk, and a short diameter of 12 mm at right
angles to it. The rachis was erect and situated distally on the
long diameter. It bore the pinnules inclined inwards in two
imbricating ranks, formed by the two rows of pinne on each
leaf, in such manner that their upper surfaces all faced towards
the axis of the trunk. |

A clearer perception of the arrangement of the pinnules
with respect to the rachis and the trank may be had if an
ordinary, expanded, young Zama leaf is considered in its
natural position on the summit of the trunk. If the pinnules
of such a leaf are folded toward the axis of the trunk in two
ranks, and these ranks are brought together side by side, the
position of the parts in the unexpanded young leaf of Oycade-
oidea will be paralleled. As just stated, it at once becomes
apparent that the rachis is farthest away from the trunk, and
that the upper side of each pinnule faces the axis of the trunk.
This is also the relative position of the parts in the unexpanded
young leaf of Zamia. Cycadeoidea thus had the direct mode
of prefoliation ehardcionieinte the ee North American
eycads Zana and Diodn.

The Cycadeoidean Leaf Structure.

While an exhaustive treatment of the anatomy of these
leaves must be deferred for the present, the features of more
immediate interest may be here described.

The exact arrangement of the pinnules and of their fibro-
vascular bundles is shown in the transverse section, figure 2
(Plate VII), and in the longitudinal section, figure 3, the
general position of these sections being indicated in figure tf.

The transverse section, figure 2, shows that, just above the
termination of the rachis, there are 20 pairs of projecting pin-
nules. Since the pinnules nearest the rachis must necessarily

G. R. Wieland—American Fossil Cycads. 307

be cut through their bases, while in a given transverse section
those farthest from the rachis must be cut through their tips
(as may be understood by referring to figure 1), this transverse
section virtually represents a series of sections such as would
_be obtained if a single pinnule were cut transversely 20 times
_at regular intervals from base to tip. Nearer the base of the
_ leaf, a transverse section cuts 34 pinnules. As there are about
_ 20 insertions of pinnules above this basal section, and appar-
_ ently no pinnules lying wholly beneath it, there are approxi-
_ mately 60 pinnules in all. The rachis and pinnular insertion
_ have not yet been studied.

The principal fact of interest shown by the longitudinal
section represented in figure 3 is the emergence of leaf tips on
the side next the axis, or rachis, of the leaf. As this section is
wholly above the termination of the rachis, this emergence

"indicates that the topmost pinnze are relatively short.

Necessarily, then, the leaf at this stage of its growth was of
_ a somewhat truncate form.

This fact, however, does not permit surmise as to the exact
leat habit of Cycadeoidea, for subsequent growth either may
_ have increased the truncation, or, because of great elonga-
_ tion of the rachis, resulted in a lanceolate leaf. Presumably
_ the study of the prefoliation and subsequent development of
_ the rachis and pinnee of living cycads, especially Zama, will
indicate what must have been the appearance of the full-grown
leaf of Cycadcoidea.

The microscopic structure of the pinnules is partially shown
in figures 4 and 5. In figure 4 is shown a transverse section
through a pinnule cutting a fibro-vascular bundle. The struc-
ture of the relatively heavy epidermis of the upper side of the
pinnule is not distinct even in this young stage, and forecasts a
dense and stiff structure. A single layer of sclerenchyma
cells of rounded section forms the hypodermis, from which a
double row of similar cells passes in confluently to the bundle,
where it divides to form a more or less complete circular row
of sclerenchyma cells enclosing the phloem and xylem. The
uniformity of this arrangement gives the transverse section of
the bundles a characteristic flask-shape. Beneath the hypo-
dermis is a distinct layer of palisade parenchyma followed by
an indistinct one grading into the rounded parenchyma cells
occupying the space between the bundles. The lower portion
of the leaf, or that beyond the bundles, is composed of well-
defined spongy parenchyma, with distinct intercellular spaces,
and is bounded by a thin epidermis, the structure of which is
not well marked. The lower surface at this stage is somewhat
glabrous, although revolute margins are apparently indicated.

308 G. R. Wieland—American Fossit Cycads.

In figure 5 is shown a longitudinal section through the pali-
sade parenchyma and bundle connection, with the hypodermis.
This section displays clearly the usual dichotomous cycadean
venation, as well as the mode of bundle increase. The pinne
are of linear-lanceolate form, and without a distinet midrib.

A comparison of these ‘structures with those of Zamia
entegrifolia shows a striking resemblance in all essential char-
acters,—a resemblance that is even more marked than is that
between the pinnee of Zamza and of some of the other existing
genera of the Zamiacew (Zamie).

The character of the ramentum surrounding these leaves
remains to be briefly mentioned. It consists of thin seales, or
hairlike forms, composed of stringy elements, and strongly
resembles the ramental cells figured by Carruthers’ and Seward?
in their respective descriptions of Bennettetes Gibsonianus and
Cycadeoidea gigantea. A transverse section of the ramentum
is shown in figure 6. The flat sides of these scales quite uni-
formly face the leaves, though as such a large portion of the
summit is occupied by a mat of relatively the same character,
it is probable that much of this hairy and scaly material is
borne on the trunk in the space inter eaee between the
petioles of the young leaves.

Yale Museum, New Haven, Conn., March 20, 1899.

EXPLANATION OF PLATE VIL.

FiGuRE 1.—Diagram of a hypothetical leaf, with a reduced number of pinnules,
showing the position of the sections represented in figures 2 and 3.
O, O', are in the plane of the transverse section, figure 2, and also
approximately mark the base of the longitudinal section, figure 3.
FIGURE 2.—Cycadeoidea ingens, Ward (type); transverse section through upper
portion of young leaf. x4. For relative position, see figure 1.
a, upper side of an inner pinnule cut near the summit; 0, fibro-
vascular bundle; c, pinnule cut near the base; d, axis of growth of
rachis. The arrows indicate approximately the base line O, 0’,
figure 3. ;
FIGURE 3.—The same specimen; longitudinal section through summit of young
leaf. x4. For relative position, see figure 1.
a, Summit of a pinnule near apex of leaf; 0, upper side of pinnule;
c, base line of section indicated in figure 1, 0, O”.
FIGURE 4.—The same specimen; transverse section through a single pinnule,
showing a fibro-vascular bundle. x 60.
a, upper side of leaf; b, hypodermis; c, sclerenchyma connection
of hypodermis with fibro-vascular bundle; d, palisade parenchyma ;
é, xylem; f, phloem; g, spongy parenchyma : h, epidermis of under
side of leaf.
FiGuRE 5,—The same specimen; longitudinal section through a pinnule just
beneath the hypodermis, cutting the palisade parenchyma. x 35.
a, forking vein; 0, palisade parenchyma.
FIGURE 6.—The same specimen; transverse section through the ramentum sur-
rounding the young leaves. x 60.

Chemistry and Physics. 309

SCIENTIFIC INTELLIGENCE.

I. CHEMISTRY AND PHYSICS.

1. On Luminous and Chemical Energy.—The radical distinc-
tion which exists between photochemical reactions which are
endothermic, in which light energy is converted into chemical
energy, and those which are exothermic, in which the light is
merely auxiliary, the greater portion of the change being effected
_by purely chemical agencies, has been recently emphasized by
_ BERTHELOT, who has pointed out that it is only the former class
of reactions which are adapted for use in the measurement of
luminous energy. He has studied a number of substances when
exposed to direct sunlight, to diffused daylight, or when kept in
the dark, with reference to the changes thereby produced. The
- substances were contained in sealed tubes and the effect of dif-
ferent wave-lengths upon them was determined by surrounding
them with colored media. Nitric acid when concentrated is
decomposed by light at ordinary temperatures, and possibly at
still lower temperatures, the reaction being similar to that which
takes place in the dark at 100°,and which yields nitrogen per-
oxide, oxygen and water. The blue rays are most effective in
_ producing the action, which is endothermic and not entirely
_ reversible. The diluted acid is not affected by light, that of
density 1°365 undergoing no change after weeks of exposure.
The nitrogen peroxide formed tends to absorb those rays which
are active in producing it. ITodic acid and oxide, while stable in
diffused daylight, are decomposed by direct sunlight, setting free
- iodine and oxygen; the reaction being analogous to that produced
by heat at 300°. It is limited by the absorptive action of the
iodine vapor and by the deposition of iodine on the surface of the
compound. Hence, while endothermic, this reaction is not suit-
able for the measurement of light energy. The action in the case
of hydrogen iodide is on the line limiting endothermic and exo-
thermic reactions, though the author’s view is that a sniall amount
of heat (0:4 calory) is evolved; though when solid iodine is set
free, it is decidedly exothermic, 6:4 calories being evolved. The
blue and violet rays are most effective, and the decomposition is
most rapid in direct sunlight, there being an intermediate pro-
. duction of hydrogen periodide (HI,?). Hydrogen bromide is not
affected by the action of light even at 100°. Nor was any action
observed with CO, either alone or mixed with hydrogen, with CO
alone or mixed with hydrogen or oxygen, with SO, or HNO,
when mixed with hydrogen. Red mercuric oxide is more rapidly
decomposed than the yellow variety. Mercurous oxide when
exposed to light yields traces of metallic mercury. No action
was noticed with lead peroxide, silver oxide, or with mercuric
chloride, bromide or iodide. Hence the author considers that the
measurement of luminous energy by transformation into chemical

310 Scientific Intelligence.

energy is to be effected by means of endothermic reactions which
are irreversible at ordinary temperatures.— C. &., cxxvii, 143-160,
July, 1898. G. F. B.

2. On the Homogeneity of Helium.—By diffusion through a
pipe stem, Ramsay and Travers separated a volume of helium
first into six fractions. The first fraction was then pumped
into the diffusion vessel, where one-half of it was diffused and
‘transferred to vessel (1). Fraction No. 2 was then added and
one-third of the mixture was diffused and transferred to vessel
(1). Fraction No. 3 was then added and one-half of the whole
diffused and transferred to vessel (2). Fraction No. 4 was then
added and one-half diffused and transferred to vessel (3); and so
on. Four complete fractionations of this sort in the case of air

‘ gave at the extreme fractions 17:37 and 22-03 per cent of oxygen
respectively. When nitrogen was thus used, prepared from
ammonium chloride and, sodium nitrite, thirty fractionations pro-
duced no variation in density. In the case of helium, repeated
fractionations of the gas from samarskite and cleveite gave a gas
of density 1°98 and of refractivity 0°1238 as the lightest fraction
and a very little pure argon as the heaviest one. Since further
diffusion did not alter the density of the lighter fraction, the
authors regard it as pure helium. “It appears therefore that
helium contains no unknown gas nor is it possible to separate it
by diffusion into any two kinds of gas; all that can be said is
that most minerals which evolve helium also evolve argon in
small quantity.” The authors had hoped to find a gas with
density 10 and atomic mass 20; since this would form with He =
4and A= 40 a triad like that of F, Cl and Mn, 19:35°5; 55.
They still regard the existence of such an element as probable.—
Proc. Roy. Soc., \xii, 316-324, January, 1898. 6. FOB.

3. On the Properties of Ozone.-—By means of liquid air, LapEN-
BURG has condensed ozonized oxygen to a liquid. Then by
allowing this to evaporate partially, he has obtained a residue
much richer in ozone, since this gas is less volatile than oxygen.
So that by repeating the operation, he obtained finally a gas
which contained 86°16 per cent of ozone and which had a density
of 1:°3698 as compared with oxygen, this density being determined
from the rate of diffusion. Hence it follows that the density of
pure ozone would be 1°456, theory assigning to it 1°5. He finds
that ozone is not as soluble in water as has been supposed, 0°01
volume only being absorbed at the ordinary temperature and _
pressure, or 0°00002 part by weight. By condensing ozonized
oxygen, allowing the oxygen to evaporate and then ascertaining
the temperature at which the residual ozone evaporated, an
attempt was made to determine the boiling point of ozone. The
oxygen volatilized at —186°, leaving 4 or 5° of an almost black
opaque liquid, the temperature of which then rose to —125°, at
which point the liquid exploded. This temperature therefore can
be taken only as a lower limit.—Ber. Berl. Chem. Ges., xxxi,
2508-2515, October, 1898. G. FYB.

Chemistry and Physics. : 311

4. On the Preparation of Alkali Nitrites.—It has been pointed
out by Divers that the common impression that in preparing
alkali nitrites by passing nitrous fumes into alkali solutions, a
method proposed by Fritsche in 1840, there is also produced
much nitrate, is erroneous. The nitrous gases are made from
nitric acid and starch or arsenous oxide, and are passed into the
concentrated solution of the hydrate or carbonate. until neutrali-
zation is effected; the vessel being loosely stoppered. The
strength of the acid and the temperature are so regulated that
the nitrogen monoxide is in excess of the dioxide. Ifthe solution
is acid, it must be boiled until neutral before exposing it to the
air. Then the solution may be freely evaporated to crystalliza-
tion. Both potassium and sodium nitrites are slightly yellow in
color and are slightly alkaline. Sodium nitrite fuses at 271° and
at 15° five parts of it requires six parts of waterto dissolve it. It
is moderately deliquescent, and its crystals are anhydrous.
Potassium nitrite is very deliquescent, is soluble in about one-
third its weight of water, and its crystals are reputed to contain
half a molecule of water.—J. Chem. Soc., |xxv, 85, January, 1899.

G. F. B.

5. On the Composition of American Petroleum.—By means of
a combined dephlegmator and regulated temperature still-head,
Youne has succeeded in preparing from American petroleum
ether a number of nearly pure hydrocarbons. In some cases the
fractional distillation was preceded by treatment with mixed
nitric and sulphuric acids, in order to remove the benzene. Evi-
dence of the presence of the following has been thus obtained:
Isopentane Bb. P. 27°95°, normal pentane B. P. 36°3°, penta-
methylene B. P. about 59°, isohexane B. P. about 61°, normal
hexane B, P. 68°95°, methylpentamethylene B. P. about 72°, ben-
zene B. P. 80°02°, hexamethylene B. P. 80°8°, isoheptane B. P.
90°3°, dimethylpentamethylene (?) B. P. about 94°, normal hep-
tane B. P. 98°4°, methylhexamethylene B. P. about i02° and
toluene B. P. 110°8°. These results show that the same classes of
hydrocarbons, parafiins, polymethylene compounds or naphthenes
and aromatic hydrocarbons are present in petroleum from all
three sources, but that Russian petroleum contains a relatively
larger amount of naphthenes and in all probability, of aromatic
hydrocarbons than Galician, and Galician a larger amount of the
same hydrocarbons than American petroleum.—J. Chem. Soc.,
Ixxili, 905-920, December, 1898. G. F. B

6. Hlectrical waves along wires.—A. SOMMERFELD has exam-
ined by means of elaborate mathematical analysis this subject,
and has calculated the rate of damping and the decrease of
velocity for the case of wires of different diameters. In the case
of a copper wire 4™™ thick with a period of 10° per second, the
electric wave velocity was 8 kilometers below the velocity of
light. With a platinum wire of ;,4,;™™- diameter, period 3°10°
per second, the velocity falls to # of the velocity of light. The
ordinary experimental method, in which two parallel wires are used,

312 Scientific Intelligence.

offers great mathematical difficulties. The author is apparently
unacquainted with the direct experimental determination of the
velocity of electric waves by Professor Trowbridge and Professor
Duane, which avoids the mathematical difficulties, and by which
a velocity closely agreeing with that of light was obtained.—
Wied. Ann., No. 2, 1899, pp. 233-290. Je.

7. Polarization and hysteresis in dielectrics—W. SCHAUFEL-
BERGER employs ellipsoids of paraffine and hard rubber, and
studies the damping of their -oscillations in the electric field of a
Kohlrausch condenser. The mathematical consideration of the
motion of the dielectric contains hypotheses drawn from analo-
gous phenomena of polarization and hysteresis in the case of iron.
It was found that the loss of energy due to hysteresis in paraffine
is 2°119 per cent and in hard rubber 63°49 per cent. ‘The latter
number is regarded as an approximation.— Wied. Ann., No. 2,
1899, pp. 307-324. Jo.

8. A new method of showing electric waves.—A simple resona-
tor devised by Rhigi consisted in a sheet of tinfoil applied to glass,
the foil being divided by a fine line. ALBERT NEUGSCHWENDER
has discovered that if a galvanometer and a battery are
placed in circuit with a silver-coated surface divided by a
fine line, that the galvanometer will show a deflection due to
a Hertzian wave if one breathes on the metallic surface or places
a wet sponge near it. The vapor converts the divided surface
into a species of coherer.— Wied. Ann., No. 2, 1899, pp. 430-432.

BA

9. Separation of long waves of heat by means of quartz
prisms.—The very important work of H. Rusens and E.
ASCHKINASS on this subject has shown that it is possible by
means Of quartz and sylvine to isolate waves of heat of great
wave-length, since these waves are not absorbed by these sub-
stances. The above authors have now entered into a discussion
of the further extension of their method. They find that intensity
of energy between wave lengths 50u and 60u is greater than the
whole energy beyond A\=60y. Since the energy between 50 and
60u is only zp aoa Of the whole amount of energy emitted by
the sources employed by them (Auer burner), it does not appear
possible to further extend their method.— Wied. Ann., No. 2,
1899, pp. 459-466. ian,

10. Theory of conduction of electricity through gases by charged
zons.—Professor J. J. THomson has reduced his observations
upon this subject to a mathematical form, and has obtained equa-
tions which embody his experimental results. He explains many
of the phenomena observed in Geissler tubes, and believes that
the most striking features of the discharge through vacuum tubes
are due to the difference in velocity between the positive and
negative ions. He points out that a study of the distribution of
intensity in Geissler.tubes supports the theory of ionization.—
Phil. Mag., March, 1899, pp. 253-268. Je De

11. The Hall phenomenon and the theory of Lorentz.—PoincaRE

Chemistry and Physics. 313

has applied Lorentz’s theory of electrical ions to the Hall phe-
nomenon and has obtained an equation which represents the elec-
tromotive force obtained by Hall. He concludes that it would be
desirable to examine whether all metals exhibit the Hall effect
when they carry currents of very high electromotive force and
whether a change of sign occurs.— Comptes Rendus, No. 6, Feb.,
1899, pp. 339-341. Jays.
12. Onthe Relation of the Surface Tension and Specific Gravity
of certain Aqueous Solutions to their State of LIonization.—Prot.
MacGregor has shown that it is possible to calculate the various
properties of solutions of potassium and sodium chloride in terms
of the state of ionization of the electrolytes contained. The rela-
tion derived, in its simplest form, is S=S,+A(1—a)n+lan.
where S is the numerical value of any property of a solution
(density, surface tension, etc.), S, that of the same property of
water under the same physical conditions, 2 the number of equiva-
lent gram-molecules per unit volume, a@ the ionization coefficient
of the electrolyte in the solution, and /and & constants, called
ionization constants, for any given property of any given electro-
lyte. Mr. E. H. Arcuzarp has carried out a similar series of
observations and calculations in the case of simple solutions of
sodium, potassium and copper sulphates, salts of more complex
molecular structure than those previously examined. He con-
cludes that the expression given represents observed values of the
surface tension and specific gravity of the solutions examined
through a range of concentration extending from 0°05 to about 0°4
or 0°5 equivalent gram-molecules per liter. Further, that it is pos-
sible by aid of the dissociation theory of electrolysis to predict
the surface tension and specific gravity of mixtures of potassium
and sodium sulphate solutions and the specific gravity of mixtures
of solutions of potassium and copper sulphates throughout nearly
the same range, within the limits of the error of observation.
And similarly, that it is possible to predict the specific gravity of
mixtures of solutions of potassium sulphate and sodium chloride.
13. On the Cause of the Absence of Coloration:in certain: Lim-
pid Natural Waters._-Professor W. Sprine has an interesting
article upon this subject in a recent number of the Bibliotheque
Universelle (Jan. 15, 1899). Recognizing the accepted fact that
pure water is distinctly b/we, he remarks that the green color
often observed is easily explained as due to the presence of more
or less muddiness, which acts with the natural blue color to give
a sensation of green. The absence of color, however, is a dif-
ferent problem, as early noted by Berzelius. The results of some
experiments have led the author to the hypothesis that the sus-
pension of minute, perhaps invisible, particles of anhydrous
ferric oxide, or hematite, would have, through their red color, the
effect of neutralizing the blue and leaving the water completely
colorless. He notes the fact that microscopic grains of hematite
are.almost always found distributed in the soil, and that terres-
trial waters have rarely the blue color which the pure liquid

Am. Jour. Sor.—Fourts Series, Vou. VII, No. 40.—ApRIL, 1899.
21

314 Scientific Intelligence.

should show. On the contrary, the water derived immediately
from snow at high altitudes, or from glaciers, contains no hema-
tite and is commonly intensely blue. The action of hydrated
ferric oxide is different, since its color is yellow, and when present
in very small amounts in the water, there is, so to speak, a struggle.
between it and the organic matter; if not overcome, the effect is
to make the water appear green to the eye.

14. A Text-Book of General Physics for the use of Colleges
and Scientific Schools ; by Cuarues 8. Hastrnes and FREDERICK
E. Beacs; pp. vill, 768. Boston, 1899 (Ginn & Company).—
The appearance of a new high orade text-book of physics is an
event of much importance to all teachers and students in this
department. The authors are to be congratulated upon the
admirable manner in which they have treated the entire subject,
particularly for their success in adequately presenting so vast a
range of topics within the limits of a single moderate-sized
volume. The work is characterized by the very careful and
thorough treatment of the fundamental conceptions of the sci-
ence; by the well-balanced relation between the different parts,
no one portion of the subject being sacrificed to another ; by the
clear and precise mathematical development of each successive
principle. At many points, moreover, though dealing with topics
often treated of, there is much that is fresh and suggestive; this
is particularly true of the chapters on Light.

The conciseness of expression and the mathematical, though
elementary, method of presentation adopted have the advantage
of brevity and accuracy, but they may leave much for the
teacher to do if his pupils are approaching the subject for the
first time. The student, however, who is properly prepared, will
find no serious difficulty, and his subsequent work in special
advanced fields will be much facilitated by the thorough ground-
ing in principles and methods which he has here obtained.

Too much praise cannot be given to the publishers for the
admirable form in which they have placed the work before the
public. The page is clear and open, with no confused crowding
of mathematical expressions, while the excellence of topoReapny,
illustrations and paper leaves nothing to be desired.

Il. GroLtogcy AND MINERALOGY.

— 1, New facts regarding Devonian Fishes. Dentition of

Devonian Ptyctodontidea ; by C. R. Eastman; Am. Nat., vol.
XXX, pp. 478-488 and 545-560. July and August, 1898. Some
new points in Dinichthyid Osteology ; by C. R. Eastman; Am.
Nat., vol. xxxii, pp. 747-768. October, 1898.—The first of these
papers is the result of study of the rich material from the State
Quarry fish-beds discovered by Prof. Calvin in the Devonian of |
Johnson County, lowa (see Ann. Rept. lowa Geol. Survey, vol.
vii, 1897); private collections made by Mr. E. E. Teller and C. E.
Monroe from the Hamilton limestone quarries of oo)

*

al ee athe Re

Geology and Mineralogy. 315

Wis.; and the Schultze collection in the Museum of Comparative
Zoology eontaining European fish remains. ‘The species described
belong to the genera Ptyctodus Pander, Rhynchodus Newberry,
Paleomylus Woodward, and ichthyodorulites are described under
the new generic names Phlyctenacanthus and Belemnacanthus.

The stratigraphic relations of the fish remains and their asso-
ciation with invertebrate fossils lead the author to agree with the
general conclusions of Stuart Weller regarding the relation of the
Milwaukee beds to the New York and lowa faunas.

“The Chimeroids give it (the hydraulic limestone of Mil-
waukee) a stamp of antiquity, suggesting that a westward migra-
tion took place during the early part of the Devonian as far as
Wisconsin, but not crossing the Mississippi Valley until the
Middle Devonian. The Milwaukee beds show the first traces of
the encroachment from the East, the Rock Island locality a some-
what later, and the state quarry limestone the last of all, with its
horde of Upper Devonian lung-fishes.”

The second paper gives a critical comparative study of the
cranial osteology of the Dinichthyids, with a particular descrip-
tion of an early type, D. pustulosus Kastman, based upon speci-
mens from the Milwaukee and Iowa localities. The author
considers this species to be the most primitive member of the
genus known. His studies lead him to regard Coccosteus and

‘Titanichthys as the extreme limits of the family Coccosteide,

with Dinichthys as an intermediate and connecting link. Speak-
ing of the Dinichthyids the author says: ‘“ But when their char-
acters shall have been fully investigated, the wide range of
variation manifested by them will be found reducible to order,
and the whole promises to constitute one of the most interesting
evolutionary series known among fossil fishes” (p. 747).

H. S. W.

2. Geological Sketch of San Clemente Island; by W. 8. T.
Smira. Extr., 18th Ann. Rept., U. 8. Geol. Survey, 1896-97,
Part I, pp. 459-496, plates lxxxiv—xevi, figs. 82-85. 1898.—San
Clemente is the southernmost island of the “ channel islands ” off
the southern coast of California. The island has been described
as a “tilted orographic block” by Lawson. The rocks are chiefly
volcanic. The oldest sedimentaries are of Miocene age. The
author recognizing this as a typical fault block, and isolated from
other land masses, the topography of which is still in its infancy,
has presented a theory as to the development of its drainage. He
also describes the structural and petrographic features of the vol-
canics, and a general description of the sedimentaries, moraines and
later deposits. H. S. W.

3. Geology of the Edwards plateau and Rio Grande plain, ete. ;
by Roserr T. Hitt and T. Wayrtanp Vaucuan. Extr., 18th
Ann. Rept., U. 8. Geol. Survey, 1896-7, Part IL; pp. 193-321,
plates xxi-lxiv, figs. 53-76. 1898.—The geological features of
the region from Austin, Texas, westerly and southwesterly
through San Antonio to the Rio Grande, are described and beauti-

316 Scventific Intellagence.

fully illustrated in this report. Although it does not attempt to be
an exhaustive paper on the statistics and engineering problems in-
cident to developing the water supplies, the geology is presented
in its relations to the occurrence of underground waters, their
sources and relations to the beds carrying them. The paper not
only presents an admirable exposition of geological structure, but
must prove of great value in guiding the inhabitants of the region
in supplying themselves with large supplies of fresh waters.
H. : w.

4, South Dakota Geological Survey, Bulletin No. eee Dy.
James EH. Topp, State Geologist; pp. 1-139, plates i-xv, figs.
1-2. 1898.—This second bulletin contains the 1st and 2d biennial
reports with accompanying papers for the years 1893 to 1896.
The papers of chief importance are accounts of the surveys of a
Section along Rapid Creek from city westward ; A reconnoisance
into northwestern South Dakota; The geology along the Bur-
lington and Missouri Railway; and Additional notes on the limits
of the main Artesian Basin, i. e., additional to those reported by
the U.S. Geological Survey i in the 17th and 18th Annual Reports.
Numerous detailed sections of the rocks and reproduced pho-
tographs of the structure are given. H. Ss. W.

5. Crater Lake, Oregon ; by J.S. Dittxr. Extr. Smithsonian
Report for 1897, pp. 369-379, plates i-xvi. 1898.—This paper,
though a reprint of one which first appeared in the National Geo-
graphic Magazine (1897),* will bear repeating, holding as it does
beautiful illustrations of a region of which the author says:

‘‘ Aside from its attractive scenic features, Crater Lake affords
one of the most interesting and instructive fields for the study of
volcanic geology to be found anywhere in the world. Considered
in all its aspects, it ranks with the Grand Cafion of the Colorado,
the Yosemite Valley, and the Falls of Niagara, and it is interest-
ing to note that a bill has been introduced in Congress to make it
a national park for the pleasure and instruction of the people.”

H. S. W.

6. International Geological Congress.—The eighth session of
the International Geological Congress will be held in Paris in
1900 between the 16th and the 28th of August. The Committee
of Organization has announced twenty-two excursions, three of
which are general; I, about the Paris Basin; II, Boulonnais
and Normandie; III, the volcanic regions of central France. The
other nineteen are special local excursions for the study of particu-
lar problems. <A Jdivret-guide will be prepared and ready for dis-
tribution the early part of 1900. M. Albert Gaudry, Paris, is the
President, and M. Charles Barrois, Lille, is General Secretary of
the Committee of Organization. H. Ss. W.

7. International Geographical Congress. —The. Geographical
Society of Berlin (Gesellschaft fiir Erdkunde zu Berlin, F. Baron
von Richthofen, Prest., George Kollen, General Secy.) invites the
friends and promoters of geography in all countries, and especially

* See also this Journal for March, 1897.

| aA

Geology and Mineralogy. 317

the members of all geographical societies and cognate scientific
bodies or institutes, to assemble at the German capital, Berlin, for
the meeting of the seventh International Geographical Congress,
to be held from Sept. 28th to Oct. 4th, 1899. Membership tickets
may be obtained by gentlemen or ladies by payment of 20 marks
to the Treasurer, H. Biitow, 90 Zimmerstrasse, Berlin S. W.
All wishing to read papers before the Congress are requested to
give notification thereof before April 1st, 1899, and to send the
manuscript ready for print not later than June Ist, 1899.
H. S. W.

8. Funafuti.—The January number of Natural Science con-
tains an interesting and popular discussion of the life and geolog-
ical features of Funafuti as a typical coral atoll, by W. J. Soutas.
Special interest attaches to this island in consequence of the
work which has been done here in boring into the coral rock.
The party sent out from England, which was directed by
Professor Sollas, unfortunately did not succeed in carrying the
boring to any considerable depth. The work, however, was con-
tinued in 1897 under the direction of Professor David of the
University of Sydney, and this time was entirely successful, the
reef being penetrated to a depth of 697 feet. Later, a third party
was sent from Sydney, which has carried the boring on to a depth
of about 1000 feet. The core obtained by the David party was
sent to England and is now in the hands of Professor Judd for in-
vestigation, the results of which will be awaited with much
interest. The general statement is made, however, that the
material brought up from the boring, of which the reef is com-
posed, presents much the same character throughout, and so far is
regarded as supporting Darwin’s theory. There are no layers of
chalky ooze, such as Murray’s hypothesis might have made prob-
able, and no trace of volcanic material was found. The later
boring beyond 700 feet has passed through a hard limestone, con-
taining many well-preserved corals. The last mentioned expedi-
tion has also carried on a boring into the bed of the lagoon toa
depth of 144 feet. After passing through 101 feet of water, the
first 80 feet below were found to consist of the calcareous alga
Halimeda, mixed with shells; the remaining 64 feet, of the same
material mixed with coral gravel.

9. Petrographical and Geological Investigations of certain
Transvaal Norites, Gabbros, and Pyroxenites; by J. A. Lro
HENDERSON (8° pamph., 56 pp., 5 pl., Dulau & Co., Pub., London,
1898).—Although this appears as a separate work, one is led to
infer that it is a thesis presented for the doctor’s degree at
Leipzig. The material upon which the investigations have been
carried is from the Stelzner collection at Freiberg and was gath-
ered by the explorers Mohr and Goerz. As may be inferred,
therefore, it is a minute and painstaking petrographical and chem-
ical investigation of a series of hand specimens determined as
diallage norite, hypersthene norite, hypersthene diallage gabbro,
quartz norite, quartz diallage norite and enstatite pyroxenite.

318 —- Serentifie Intelligence.

In addition some gneisses and other rocks are described. Among
them is a syenite and syenite porphyry which are shown to have —
anorthoclase (soda microcline) as the feldspar, and the author
believes this to be a sufficient ground for the erection of a new
rock group and proposes the names of hatherlite and pilandite for
them respectively. Probably the majority of petrographers will
hardly agree to this, since as commonly understood the syenite
family is one composed essentially of alkaline feldspars in general
and not one particular variety of them. As a matter of fact, the
pure potash feldspar is of very rare occurrence as a rock-ingre-
dient and the great majority of the so-called orthoclases contain
the albite molecule to a greater or less extent. Moreover since
the analyses of the feldspars of the rocks mentioned show that
they contain a very considerable quantity of lime, it appears ques-
tionable whether they should not be classed as abnormal oligo-
clases rather than anorthoclase (soda microcline) ; they are clearly
not typical examples of the latter.

Very full and painstaking investigations have been made in
this work of the augitic constituents of the rocks mentioned
which have been separated and analyzed, but it seems rather
curious from modern standpoints that with so much chemical
investigation of the constituents no mass analyses of the rocks
except In one case have been made. This is a rock consisting of
enstatite crystals including 4°5 per cent of anorthite plates. The
analysis gave : |
SiO. On AICO, We,0.’ vee Mg CaO H,0°, Sp2Gi=3

55°23 0°44 2°08 3°94 6°25 29°29 1°68 LA == Sr Ons
It is an enstatite pyroxenite from the Central Marico District.
. ine Vv. JE

10. On the Occurrence of Corundum.—Two recent papers on
the occurrence of corundum in widely separated parts of the
world deserve to be mentioned. One of these, by T. H. Hottanp,*
discusses in detail the various corundum localities, including the
gem varieties, ruby and sapphire, in India and the adjacent
countries. The second paper, by W. G. Mirzer,ft describes the
corundum and associated minerals of Eastern Ontario.

The most interesting point brought out in both papers is the
fact that corundum occurs so often associated with igneous rocks ;
this fact has also been recently established in other regions. Mr.
Holland accepts the fact that in certain cases it may have a sec-
ondary origin, as shown by Professor Judd for the ruby, but for
the most part he regards it as an original constituent, though
often due to the local excess of alumina. ‘The Indian localities
are divided into those in which the corundum occurs with basic
rocks and these in which its associates are siliceous, that 1s, aczd.
In the former class pyroxene is the predominating constituent,

* A Manual of the Geology of India: Economic Geology by the late Professor
V. Ball. Second edition revised in parts. Part I, Corundum, by T. H. Holland;

pp. 79, Calcutta, 1898.
+ Report of the Bureau of Mines, vol. vii, third part, pp. 207-265, Toronto, 1898.

Miscellaneous Intelligence. 319

together with some member of the spinel group. These occur-
rences are divided into three groups. Inthe first, the ferruginous
group, the pyroxene is a highly ferriferous enstatite and the spinel
_ either hercynite, or hercynite with magnetite ; in this group
_ ilmenite may largely replace the corundum. In the second, the
Serro-magnesian group, ferrous oxide is largely replaced by mag-
nesia, a less ferriferous enstatite and pleonaste, also sometimes
chrysolite, being the type minerals. In the third, or magnesian
group, iron compounds occur in very small quantity, the associ-
ated minerals being magnesian forms of chrysolite, also (secondary)
talc, dolomite, magnesite.

In the second class, the corundum is associated with acid

rocks. This is well illustrated by the sapphire of Kashmir, which
occurs in granite; here the granite (pegmatite) forms veins in
gneiss and carries with the sapphire, tourmaline, euclase, cyanite,
etc. ; green tourmaline, spodumene, cookeite and prehnite are
also found. The gneiss is coarse and schistose, with white feldspar,
black mica and also garnets.

The corundum of Ontario, as described by Mr. Miller, belongs
chiefly to dikes of syenite and quartz pegmatite intersecting the
Laurentian gneiss. It is found more abundantly in ordinary
syenite than in nepheline syenite, but in the latter the crysta's
are better formed. It is also found in eastern Ontario in crys-
talline limestone, as in Burma, where it is regarded as of second-
ary origin. Mr. Miller’s paper is accompanied by numerous
interesting plates illustrating the actual occurrence of the min-
eral. Other mineralogical points are also brought out, as the
occurrence of nickel-bearing magnetite, also of a mineral of the
columbite type (G.=5'40), in Renfrew County.

11. Minerals in Rock Sections : The practical methods of iden-
tifying minerals in rock sections with the microscope. Especially
arranged for students in technical and scientific schools. By La
McItvaine Lvuqurer; pp. 117. New York, 1898 (D. Van
Nostrand Company).—This subject has been already treated of
by a number of authors, but there still remains place for the
present volume. It gives in the opening chapters a concise sum-
mary of the optical principles, methods and instruments, and
following a statement of the characters of minerals arranged with
system and clearness. A chapter is also added on the methods of
making sections and one on chemical tests. An optical scheme
is added in the appendix.

12. New Mineral Names.—Lacoriouitre. In connection with
his important work in mineral synthesis and the crystallization of
minerals from a molten magma, Morozewicz has obtained iso-
metric crystals or crystalline grains of a mineral having the
composition (Na,Ca),Al,(SiO,),, in which Na,:Ca=3:2. This
corresponds to a sodium-grossularite, a type of garnet not yet
noted in nature but which has peculiar interest as connecting the
garnets with the minerals of the Sodalite Group. This compound
is named after Prof. A. Lagorio.— Min. petr. Mitth., xviii, 147,
1898.

320 Scientific Intelligence.

THALENITE. A new yttrium silicate, having the composition
H,Y,Si,0,,. It occurs in tabular monoclinic crystals of a flesh-
red color. Hardness = 6'5; specific gravity = 4°229. Found in
Osterby in Dalecarlia, Sweden, and named by Benedicks after
Prof. R. Thalén.— G. FGr. Forh., xx, 308, 1898.

Kryprire. A name given by Lacroix to the calcium carbonate
in the form of pisolites from Carlsbad and Hammam-Meskoutine in
Algeria, which has formerly been referred to aragonite. The spe-
cific gravity varies from 2°58 to 2°70, or less than that of calcite.
Optically the birefringence is 0:020. In parallel polarized light
a distorted black cross is noted, while portions give a positive
black cross in converging light. Heated to low redness, the
pisolites decrepitate and are finally transformed into calcite; the
name given refers to this fact.— Comptes Nendus, cxxvi, 602, 1898.

IIL. MiIscELLANEOUS SCIENTIFIC INTELLIGENCE.

1. The Benjamin Apthorp Gould Fund.—A_ circular has
recently been issued by the Directors of the Gould Fund, Pro-
fessors Lewis Boss, Seth C. Chandler and Asaph Hall, stating
that a sufficient available income has now accrued to warrant
beginning its distribution, and that they are prepared to receive
and consider applications for appropriations. It will be remem-
bered that the fund was established in November, 1897, when the
sum of twenty thousand dollars was given to the National
Academy of Sciences, as trustee, in memory of the father of the
donor, Miss Alice Bache Gould, the income to be used to assist
the prosecution of researches in astronomy.

in order to guide those desiring to secure appropriations from
the fund, the directors make the tollowing statement :

The objects of the institution are, first, to advance the science
of Astronomy ; secondly, to honor the memory of Dr. Gould by
ensuring that his power to accomplish scientific work shali not
end with his death. In recognition of the fact that during Dr.
Gould’s lifetime his patriotic feeling and ambition to promote the
progress of his chosen science were closely associated, it is pre-
ferred that the fund should be used primarily for the benefit of
investigators in his own country or of his own nationality. But
it is further recognized both by the donor and the directors that
sometimes the best possible service to American science is the

‘maintenance of close communion between the scientific men of -

Europe and of America, and that, therefore, even while acting in
the spirit of the above restriction, it may occasionally be best to
apply the money to the aid of a foreign investigator working
abroad. The wish was also expressed by the donor that in all
cases work in the astronomy of precision should be given the
preference over any work in astrophysics, both because of Dr.
Gould’s especial predilection and because of the present existence
of generous endowments for astrophysics.

Finally, the Benjamin Apthorp Gould Fund is intended for the

Miscellaneous Intelligence. 321

advancement and not for the diffusion of scientific knowledge,
and is to be used to defray the actual expenses of investigation,
rather than for the personal support of the investigator during
the time of his researches, without absolutely excluding the latter
use under the most exceptional circumstances.

Application for appropriations from the income of this fund
may be made informally by letter to any of the directors, stating
the amount desired, the nature of the proposed investigation, and
the manner in which the appropriation is to be expended. If
favorably considered, a blank for formal application will be for-
warded for signature, with the rules adopted by the directors for
the administration of the fund, and to which the applicant will be
expected to subscribe.

2. Annual Report of the Board of Regents of the Smithsonian
Institution for the year ending June 30, 1896. Report of the
U.S. National Museum ; pp. xxiv, 1107. Washington, 1898.—
The Report of the United States National Museum for the year
ending June 30th, 1896, has recently been issued. It includes the
report of the Assistant Secretary, the late Dr. G. Brown Goode,
which gives, in addition to matters of current interest, an outline
of the history of the Museum, showing how it has grown to its
present dimensions. A series of Appendices relating to Museum
matters follow.

Part II of the volume, covering nearly 750 pages, includes a
paper entitled: An account of the United States National Museum
by Frederick W. True (reprinted); also aseries of ethnographical
papers elaborately illustrated with many text figures and full
page plates. The first of these, by Thomas Wilson, discusses the
origin of art as manifested in the works of pre-historic man;
those not acquainted with the subject will be surprised to see
what a variety of objects is here described and illustrated and
how great interest attaches to them. The other articles are on
Chess and Playing Cards by Stewart Culin ; Biblical Antiquities
by Cyrus Adler and I. M. Casanowicz; The Lamps of the Eskimo
by Walter Hough.

3. The Second Washington Catalogue of Stars: together with
the Annual Results upon which it is based. The whole derived
from Observations made at the U. 8. Naval Observatory with an
8°5 inch transit circle, during the years 1866 to 1891, and reduced
to the epoch of 1875:0. Prepared under the direction of Joun R.
Eastman, Professor of Mathematics, U. 8S. N., Washington Observa-
tions for 1892, Appendix I, pp. Ixiii, 287, Washington, 1898.—This
large volume of star observations, as stated in the preface by the
Director of the Observatory, Professor Harkness, has absorbed the
labors of about two-thirds of the Observatory staff for more than
thirty years, during the greater part of which time Professor East-
man was in immediate charge. His personal work included 17,334
observations of stars, out of nearly 73,000; and nearly 40,000
were made under his immediate direction. Professor Harkness
adds that the credit and responsibility for the methods of discuss-

322 Scientific Intelligence.

ing the observations, for the values of the systematic corrections
applied in the formation of the final places, and for the arrange-
ment of the volume rest with Professor Eastman.

4. Zoological Results based on material from New Britain,
New Guinea, Loyalty Island and elsewhere, collected during the
years 1895-1897 ; by ArtHuR Wi.iEy, Cambridge, England,
Part II, pp. 121-206, plates xii to xxiil.—The first part of this
important series of memoirs was noticed in the January number of
this Journal (p. 79). The present part contains the following
seven papers: 7, Report on the specimens of the genus Millepora,
by S. J. Hickson, pl. xii-xvi; 8, On the Echinoderms (other than
the Holothurians) by F. Jeffrey Bell, pl. xvii; 9, Holothurians,
by F. P. Bedford, pl. xviii; 10, On the Sipunculoidea, by A. E.
Shipley, pl. xviii; 11, On the Solitary Corals, and 12, On the
postembryonic development of Cycloseris, by J. Stanley Gardi-
ner, pl. xix, xx; 138, On a collection of Earthworms, by F. E.
Beddard, pl. xxi; 14, On Gorgonacea, by Isa L. Hiles, pl. xxii,
xxiii, Part III is stated to be now in press.

5. A Select Bibliography of Chemistry, 1492-1897 ; by
Henry Carrineton Bouton. First Supplement; pp. ix, 489.
Washington, 1899. (Smithsonian Miscellaneous Contributions,
1170.)— Workers in science owe much to the careful and pains-
taking labors of those who prepare the bibliographies in the
different departments. One of the most complete of these is the
Bibliography of Chemistry, 1492-1892, by Professor Bolton,
which was issued in 1892. A First Supplement to this work has
now been issued, bringing the literature of the subject down to
the close of 1897. This is a closely printed volume of nearly
five hundred pages, which fact alone gives evidence of the large
amount of material brought together. The total number of
titles is 5554 in twenty-five different languages; these titles
are classified under the following seven heads: Bibliography ;
Dictionaries and Tables; History of Chemistry ; Biography ;
Chemistry, Pure and Applied; Periodicals.

6. Biological Laboratory.—The Biological Laboratory of the
Brooklyn Institute of Arts and Sciences will be located for its
tenth season at Cold Spring Harbor, Long Island, during the
months of July and August, 1899. The director of the laboratory
is Charles B. Davenport, Ph.D. of Harvard University. A circu-
lar giving the names of instructors, courses of study offered,
equipment and other details has recently been issued, and copies
may be obtained from Prof. F. W. Hooper, 502 Fulton st.,
Brooklyn.

a Fe eaes

A new and small find of this rare and _ beautiful
Uranium mineral. recently made in Colorado, has all come
to us. The specimens are even more attractive than
those in the little lot secured a year ago. which received
so warm a welcome from collectors, and. include several
large size specimens at $1.00 to 36.00, and an exccllent
assortment of smaller cabinet sizes at lic. to 31.00.

Johannite from Colorado. The best specimens we
have ever had or seen, 35e. to $1.50.

Uraninite, large masses of rich. velvety-black min-
eral, of exceptional purity, also from Colorado, 50¢. to
: $5.00.

- Halite! An order given long since has just been executed for a lot of the
matchless, transparent Halite crystals from Stassfurt, Germany, with movable
bubbles. They range in size from 2x2x14 inches to 5x4x2 inches, and in price
from 50c. to $1. 50. Some of the erystals have one large bubble, others several
- small ones. We have also an excellent assortment of specimens without. bubbles,
both in transparent crystals and cleavages, which we are seliing at very low
‘prices, 10c. to 50c. As illustrations of cubical cleavage 10 mineral equals this

_ German Halite.

_ Gorgeous Labradorite. We are having a large number of specimens of the
~ Labrador Labradorite polished and a number of strikingly fine pieces are now in
_ stock and others will shortly be completed- Our prices for the specimens in the
+ rough are the lowest ever known, while the polishing will be done at the rate of
ie Seven cents per square inch.

GRAND ENGLISH FLUORITES AND BARITES.

a An importation, just arrived. brings us a few of the best Fluorites ever found.
_ (a) Flattened or elongated cubes, quite perfect and simple (not twins), 50c. to $1.50.
' (b) Purple or violet crystals, twins and groups. the cubes ranging from one inch
_ to 52 inches, 25c. to $8.00. (c) A very few Fluorite groups sprinkled. over with
~~ Quartz crystals rivaling Herkimer County in clearness and brilliancy, 35c. to $1.50.
(d) Also 220 small groups and loose twins of bright green Fluorite at 5c. to 25c.
_ (e) Elongated prisms of Barite of beautiful amber-yellow, a new find, 25c. to
$1.00. (f) Tabular brown crystals of a new type, showing excellent phantoms;
_ loose erystals, 10c. to $2.00; groups, 25c. to $7.50. (g) Choice white, tabular
_erystals and groups of Barite, 25¢. to $3.50. (h) An excellent lot of the brilliant
_ Pallafiat Calcites, 15c to $2.00. (i) A few of the popular Yellow Phantom
_ Calcites from the Stank Mine, 25c. to $2.00. (j) A large lot of groups of red
_ Bigrigg Mine Calcites, 10c. to $1.50, amd several hundred loose crystals.
_ (k) 160 beautiful sparkling crystallized Hematite and Quartz, at 5c. to 35c.
_ (i) 42 attractive Sphalerites with Dolomite, 10c. to $1.00.
: Ohio Selenite Crystals of extra large size and fine quality, 15c. and 20c.
_ each; smaller or less perfect crystals, 5c. to 10¢. each; inferior crystals, 10 for
- 10c. to 4 for 10c. We have never before seen so fine or so large an assortment.
Japanese Quartz Twins, extra large and fine, $2 50 to $17.50, one 44 inch
' erystal on matrix $65.00.
_ Japanese Stibnites, choice, terminated, loose crystals, 25c. to $5.00.

Rare Tellurium Minerals, recently received from Colorado, and each
_ specimen individually identified; Calaverite, 25c., 35¢., $1.50 and $2.00. Al-
a taite, $2.25. Coloradoite, $5. 00. Petzite, 25c to $3.00 ; Hessite, 25c. to
$1.00. Sylvanite, 35c. to $2.50. Tellurium, 10c. to $4.00, a few of the
J Specimens’ well crystallized.
Colorado Gothite, in groups of radiated crystals, occasionally terminated, 25c.
sto $1.25.
_ Other Recent Additions include a small lot of Wonderfully Strong
_ Lodestone, a large lot of showy Montana Aurichalcites, splendid Arkansas
4 Wavellite and Variscite, loose crystals of Aragonite and Titanite, groups of large
a _ Martite crystals, a most attractive lot of Chalcosiderites, etc., etc.

% —-124- page ILLUSTRATED CATALOGUE, 25c, in paper; 50c. in cloth.
_ 44-page PRICE-LISTS, also BULLETINS and CIRCULARS FREE.

GEO, L. ENGLISH & CO., Mineralogists,

REMOVED TO :

.
e po.
er
a

Oe

URACONITE

Ll

7 - 7. 29 a
ivy Jax :
or. se} Fe
kh n

_ aA 2 ni
rte =

ys vets
Sonate

sh Ae
gnaw,

os
So St

é is \
es Cin deerme" .
foe mate S WPS od ae te
ae

&
ri 2

5

ede
The

> & &

a) eH
a ?
4

- —

et
cy

Sty. 4 ah
A ae
2 kos

*
PA
wt

CONTENTS.

Arr, XXV.—Glacial Lakes Newberry, Warren and Dana,
in Central New York; by H. L. Fartronitp. (With
Plate VI) ! :

XXVI.—Rapid Method for the Determination of the
Amount of Soluble Mineral Matter in a Soil; by a. Hee
MANS (2 0 2obl. 222 Se ee

XXVIL—New Type of Telescope Objective especially
adapted for Spectroscopic Use; by C. 8S. Hasrrnes -_--

XXVIII.—Phenoerysts of Intrusive Igneous Rocks ; by L.

- V. Pirsson

XXIX.—Occurrence, Origin and Chemical Composition of
Chromite; by J. H. Pratr.

XXX.—Influence of Hydrochloric Acid in Titrations by
Sodium Thiosulphate, with special reference to the Hsti-
mation of Selenious Acid; by J. T. Norton, Jr

XX XI.—Rock-forming Biotites and Amphiboles; by H. W.

XX XII.—One little known and one hitherto unknown species
of Saurocephalus; by O. P. Hay
XXXIII.—Some American Fossil Cycads. Part II; by G.
OW. Wievann.:. (With Plate’ V I.) 3) °s00ee ee

SCIENTIFIC INTELLIGENCE.

Chemistry and Physics—Luminous and Chemical Energy, BERTHELOT,
Homogeneity of Helium, Ramsay and TRAVERS: Properties of Ozone, LADEN- |
BURG, 310.—Preparation of Alkali Nitrites, Divers: Composition of American f{
Petroleum, Youne: Electrical waves along wires, A. SOMMERFELD, 311.—
Polarization and hysteresis in dielectrics, W. SCHAUFELBERGER: New method —
of showing electric waves, A. NEUGSCHWENDER: Separation of Jong waves of —
heat by means of quartz prisms, H. RuBENS and H. ASCHKINASS: Theory of —
conduction of electricity through gases by charged ions, J. J. Taomson: Hall —
phenomenon and the theory of Lorentz. Porncar#, 312.—Relation of the Sur- —
face Tension and Specific Gravity of certain Aqueous Solutions to their State of ~
Ionization, KE. H. ARCUBALD: Cause of the Absence of Coloration in certain —
Limpid Natural Waters, W. Spring, 313.—Text-Book of General Physics for —
the use of Colleges and Scientific Schools, C. S. Hastrnas and F. E. Braca, 314, —

Geology and Mineralogy—New facts regarding Devonian Fishes, 0. R, HAst- —
MAN, 314.—Geological Sketch of San Clemente, Island, W. 8. T. SMire: Geol- |
ogy of the Edwards plateau and Rio Grande plain, ete., R. T. How and TW,
VAUGHAN, 315.—South Dakota Geological Survey, Bulletin No.'2; J. i Topps ae
Crater Lake, Oregon, J. 8S. DILLER: International Geological Congress: Inter- |

national Geographical Congress, 316.—Funafuti, W. J. Sonias: Petrograph-
ical and Geological Investigations of certain Transvaal Norites, Gabbros, and
Pyroxenites, J. SAY L. HENDERSON, 317.—Occurrence of Corundum, POE HoL- b:
LAND and W. G. Miniter, 318.—Minerals in Rock Sections, L. McL. LUQUER :
New Mineral Names, 319.

Miscellaneous Scientific Intelligence—Benjamin Apthorp Gould Fund, 320.—Ar
nual Report of the Board of Regents of the Smithsonian Institution for th
year ending June 30, 1896: Second Washington Catalogue of Stars, 321,
Zoological Results based on material from New Britain, New Guinea, Loyalty |
Island and elsewhere, collected during the years 1895-97, A. Witney: Select |
Bibliography of Chemistry, 1492-1897, H. C. Bouton: Biolegiea Laboratory,
299 ‘

Vad.

Talcott,

S. Geol. Survey.

i AMERICAN
JOURNAL OF SCIENCE.

Epitrorn: EDWARD S. DANA.

ASSOCIATE EDITORS

Prorzssors GEO. L. GOODALE, JOHN TROWBRIDGE,
_ H. P. BOWDITCH anv W. G. FARLOW, or Camarines,

| Prorzssors A. E. VERRILIL, HENRY 8. WILLIAMS, anv
. I. V. PIRSSON, or New Haven,

r

(
hee
|

Prorzssor GEORGE F. BARKER, or Puimapetruia, e
Proressor H. A. ROWLAND, or Batrimorge, Rogen
Mr. J. S. DILLER, or Wasuincron. eg

FOURTH SERIES,
VOL. VII—-[WHOLE NUMBER, CLVII.]

No. 41.—MAY, 1899. : ee

WITH PLATES VIII-X.

NEW HAVEN, CONNECTICUT.
1899.

TUTTLE, MOREHOUSE & TAYLOR, PRINTERS, 125 TEMPLE STREET.

plished monthly. Six dollars per year (postage prepaid). $6.40 to
subscribers of countries in the Postal Union. Remittances should

2 a

QUEENSLAND i

- PRECIOUS OPAL

Gorgeous Matrix Specimens. at Unprecedent— z
edly Low. Prices. : ay ee

A large collection of this beautiful and popular stone, purchased by our tte >a
tralian collector, was opened for sale May Ist. A description of so well known a —
gem seems almost unnecessary, but the purchase of a specimen will explain the
use of superlatives in referring to it. Of few things is it as easy to say “finest,” =~
as of the happily named Noble Opal. An exquisite play of delicate colors,or ~~
perhaps bolder and grander flashes of varied lights, spread showily over a broad
‘surface of rock, have won a reputation for this gem among all others. The ~~
prominent colors are green and blue, with splashes of red and purple. Often the ~
opal has considerable “depth, yielding gem material of the first quality. ‘There is 1" 4
no ‘‘ washed out” look about such specimens. They present a natural display of.
combined and brilliant color, in varied changes, that are alike ee despair of artist
and counterfeiter. 4

The Queensland opals come from mines in the Bulla Creek diamiet which have
for several years produeed a large proportion of the world’s supply. They oceur
in a brown limonite matrix which is often of a flinty texture, serving asasombre ~~
but excellent setting. Our present prices are,most reasonable, barely covering
the actual gem value. The best generally range from 1 to 2 inches diameter,
and are priced at $2.00 to $8.00. A few at $10.00 to $15.00, including polished
specimens of unusual beauty. Gem fragments and pieces of matrix showmg

_ charming bits of color, as low as 1ldc. to $1.50. They are infinitely superiortothe —

Mexican, and should be in every collection. A good stock of cut opals, 50c. to — a

$8.00 per earat.
CROCOITE.

NEW CONSIGNMENT. 50 PER CENT. REDUCTION.

Our introduction during the past year, of this rare and showy crystallization, —
has met with phenomenal success. The first stock of good specimens was about. —
exhausted, but is replenished by a small series just received. Although an old
find they were but recently bought in by our collector, and at such terms as
enable us to cut former rates in half. Gorgeous orange-red masses of crystals;
ferromanganese with lustrous and perfectly terminated erystals of varied habits,
at $1.00 to $4.00. ‘Several at $5.00 to $8.00. Detached crystals and small
matrix specimens, 15c. to 75c. Pure erystal fragments, 50c. per oz. ~

Mesolite, Phacolite, Phillipsite, Ferrocalcite, Aragonite and. ae .
minerals from the Collingwood Quarries. near Melbourne, reached us in the same
consignment. See Catalogue. vs

Many other New Arrivals are mentioned in our 16 pp. Z lustrated Supple- te
ment, including Crystal and Pound Material addenda lists, New Species, Traut-
wine Collection, ete.. etc., mailed free.

Illustrated Collection Catalogue, 64 pp. Prospectors and educational .
sets. .Free.

Illustrated Catalogue of Cabinet Specimens, 126 pp. Free.

Is” MINERALS CS. ee

Dr. A. KE. i. FOOTE,
WARREN M. FOOTE, Manager.
Minerals for Scientific and Educational Purposes.

1317 Arch Street, Philadelphia, Pa., U. > A.

ESTABLISHED 1876.

beacs

i : v - 7 ‘ wee ; 4
eer tee Me ee ee ae ‘i
Taree, OPT Sh Renee ec ee a Lene

»
ete

=

ee Teinntete eres

THE

AMERICAN JOURNAL OF SCIENCE

[FOURTH SERIES. ]

Arr. XXXIV.—On some Experiments with Endothermic
. Gases; by W. G. Mrxter. .

| [Contributions from the Sheffield Laboratory of Yale University. ]

THESE experiments were made in U-tubes of 30 to 50™ in
length and for high pressures the form shown in fig. 1 was

employed. The part above the electrodes @ a was narrowed
for convenience in opening the tube and measuring small
residual quantities of gases. The other end was connected by
means of a small copper tube with a hydraulic pump. At 6

Am. Jour. Sor.—Fourts Series, Vou. VII, No. 41.—May, 1899.

324 Mixter—Kaperiments with Endothermic Gases.

was a thick wrapping of cotton wicking to protect the rubber
stopper from the wire used to hold it in place. Such a connec-
tion answered for a pressure of 20 atmospheres by the pump
and withstood explosions that burst the other limb of the U-
tube. For filling the apparatus a little mercury was poured in
to form a trap at the bend and the air was displaced by passing
in a liter or more of gas through a small rubber tube reaching
to the closed end; then more mereury was added, the stopper
adjusted, the apparatus placed in a tall glass jar to prevent
danger from flying glass and to catch the mercury in case the -
tube broke. The desired pressure was obtained by foreing in
water with a pump. For pressures of three atmospheres or less
a reservoir of mercury connected by means of a rubber tube
was convenient, the heighth of the mercury in the reservoir
above its surface in the U-tube being the measure of the pres-
sure. For heating gases by a glowing wire the end with the
electrodes in fig. 1 was modified as indicated in fig. 2. To the
heavy copper wires passing through the rubber stopper was
soldered a platinum wire which was made to glow or even
melt by a battery current.

Acetylene.

Berthellot and Vieille* have investigated the explosion of
acetylene, and hence it is only necessary to state
that-my results, cbtained before their paper was
published, accord with theirs and that they are
briefly given in order to show that the other
endothermic gases tested do not explode in the
same apparatus and under the same conditions.
At a pressure of two atmospheres when an elec-
tric spark was passed or a platinum Wire was made
to glow in the gas it sometimes did not explode,
while at a pressure of three or more atmospheres
explosions always occurred. If, however, the
sparks are small acetylene does not explode even
at high pressures. A number of tests were made
in a tube shown in part in fig. 3, containing acet-
ylene under various pressures up to 20 atmos-
pheres; when the inner and outer wires were
connected with an induction coil in action, sparks
visible in strong daylight passed between the glass
and the inner wire. This form of discharge did
not cause explosion. When it continued for some
time a reddish deposit appeared. This was insol-
uble in ordinary solvents, and exploded when heated, leaving a
black residue. It was not further investigated. The presence

* Compt. Rend., exxiii, 523.

Mixter— Experiments with Endothermic Gases. 325

_ of an indifferent gas checks the explosion of acetylene, as the
following show: A dry mixture of 1 volume of acetylene and
2 volumes of hydrogen at a pressure of 5 atmospheres was not
_ exploded by sparking; carbon separated slowly but there was
no cloud of soot. The result was the same with equal volumes
of the two gases at 6 atmospheres, while with a mixture of 2 of
acetylene and 1 volume of hydrogen at 5 atmospheres, the ex-
_ plosion promptly followed the electric discharge. The spark
_ used was 2° in air and was passed between electrodes 3™™
apart in the gas.

The following experiment illustrates beautifully and simply
_ the endothermic character of acetylene and also shows that a
_ dull red heat is sufficient to decompose it. A slow current of
_ the gas is passed through the tube gq, fig. 4, into the larger tube,
which is heated as shown bya Bunsen burner. When the
_ part of the onter tube about the end of a is heated to a faint
_ glow there is a slight puff and then the gas issuing from the
_ end of @ glows or rather the carbon, separating beyond the
end of the little tube, glows. It is a jet of gas filled with a
cloud of lampblack glowing with the heat resulting from the
_ dissociation of the acetylene. In order to obtain the result the
flow of the gas must be carefully regulated. If too fast, the
gas does not attain the temperature of decomposition, and if
too slow it does not glow.

Cyanogen.

Experiment 1.—The U-tube, fig. 1, was charged with cya-
_ nogen made from dry mercuric cyanide. The gas was nearly
_ free from air or other permanent gases, as it was found that
_ 100° left only a small volume of gas when liquefied by pres-
sure. When induction sparks were passed through the vapor
_ of cyanogen at a pressure of about five atmospheres, at which
a portion of the compound was liquid, carbon separated at
once and connected the electrodes. There was no explosion
and no decomposition except in the path of the spark. <A
primary current from two large storage cells was next passed,
the carbon filament connecting the platinum wires in the tube

326 = =Miater—Hxperiments with Endothermic Gases.

serving as a conductor. The result was a brilliant violet light
and the separation of a little more carbon. ‘The primary ecur-
rent had, however, broken the filament so that on connecting
again with a coil in action sparks passed, but soon ceased. The
primary current then gave the same result as before. The
amount of cyanogen decomposed was small, for of the 100%
taken only 1° was not absorbed by potassium hydroxide after
the experiment.

Lixperement 2.—Cyanogen was condensed in the apparatus
fig. 2 until a portion liquefied, and the wire was heated white
hot by a current from a battery. <A slight fog appeared but
there was no separation of carbon throughout the gas, and
there was no explosion. The result was the same at lower
pressures. Finally the platinum wire was melted in the gas at
the pressure of liquefaction, but there was no explosion nor
any considerable decomposition. A thin coating of carbon had
formed on the wire.

Nitrous Oxide.

Experiment 1.—The apparatus, fig. 1, was filled with 72° of
pure dry nitrous oxide, and the gas was condensed to 6% and
sparked. There was no explosion. Faint red fumes appeared.
On lowering the pressure after the sparking to that of the
atmosphere, the gas showed increased volume but not 50 per
cent, as would have been the case had the dissociation been
complete. BABES.

Faperiment 2.—100° of gas condensed to 5“ and finally to
3° were sparked for sometime. There was no explosion and the
volume at the close of the experiment was less than 150°.

Experiment 3.—T0° of nitrous oxide were condensed to 5%
and sparked. The result was the same as in the preceding
experiments.

Experiment 4.—100°% of nitrous oxide condensed to 10%
were sparked. There was no explosion.

In the above tests the platinum electrodes were 4™™ apart
and the spark used was fairly strong and had a length in air
Ole oars 3

Nitric Oxide.

Experiment 1.—Dry nitric oxide was sparked at pressures
varying from 12 to 20 atmospheres. There was no explosion,
but the gas became reddish. |

Lixperiment 2.—100° of dry nitric oxide were condensed to
10°. ‘The spark and electrodes were as described under nitrous
oxide. The first spark through the gas produced a small flame
between the electrodes. The sparking was continued only a

Minter— Partial non-explosive Combination, etc. 327
few seconds. The gas remaining was nearly all absorbed by a
solution of ferrous sulphate, showing that only a small portion
of the nitric oxide had been decomposed.

Experiment 3.—50° of gas were condensed to 6° and sparked
ashort time. After treatment with ferrous sulphate only 0°5°
remained. 7

The deportment of nitrous and nitric oxide in the above
_ tests is much the same as in the well known experiments of

_ sparking these gases at ordinary pressure.
| The results of the experiments described in this article will
be discussed in the next paper. ihe

Art. XX XV.—On a Hypothesis to explain the partial non-
explosive Combination of Hxplosive Gases and Gaseous Mia-
tures; by W. G. MIxtTEr. 7

[Contributions from the Sheffield Laboratory of Yale University. |

Mucu has been done in investigations of the conditions
under which gases combine gradually or with explosive vio-
lence. The velocity of explosive waves has been determined,
and the influences of pressure, and of excess of one gas or
dilution with an indifferent one, have been studied. Dixon*
states, ‘The explosive wave is propagated not only by burnt
molecules, but also by those of the heated but yet unburnt
molecules.” But so far as the writer is aware, no hypothesis
has been advanced to account for the fact that explosive
mixtures will not explode at low pressures, and that weak
electric sparks may not cause explosions, where strong sparks
do. Victor Meyer, with the aid of his pupils, has made a large
number of experiments with explosive mixtures at different
temperatures and pressures, but he has offered no explanation
of the phenomenon of slow combination in such mixtures. As
the union of hydrogen and oxygen has been more thoroughly
studied than that of other gases, the deportment of detonating
gas, by which is understood equivalent amounts of hydrogen
and oxygen, under various conditions, affords the best basis
for discussion. The behavior of certain other gases, and the
observations of different experimenters, will also be considered.

Dixon+ found that detonating gas at a pressure of 70™™ does
not explode when an electric spark is passed through it, and

* Chem. News, Ixvii, 39. + Trans, Roy. Soc., 1884, 634.

328 = Miater—Partial non-explosive Combination of

L. Meyer and Seubert* observed at the ; same pressure that
three-fourths of the gas combined. The writer noticedt
that weak sparks did not cause an explosion of the gas. This
is not peculiar to detonating gas, as Dixont states that “The
explosion of cyanogen and oxygen depends solely on the nature
of the spark. A strong spark .causes the mixture to explode
violently, whether wet or dry. A weak spark may be passed ~
through the mixture, wet or dry, without apparent effect.”

Emichg has recently studied the inflammability of thin layers
of explosive mixtures of gases, and has found that the length
of the spark required for ignition bears an inverse relation to
the density, and that the addition of oxygen to detonating gas
till it is to the hydrogen as 1:1 increased the sensitiveness of
the mixture to sparks; an analogous fact obtains for an explo-
sive mixture of hy drogen and chlorine. On the contrary,
equivalent mixtures of carbonic oxide and oxygen are more
sensitive than those containing an excess of either.

The glow discharge of electricity produces gradual combina-
tion of explosive mixtures.| Meyer and Raum] experimented
with detonating gas in sealed bulbs, and observed no percepti-
ble combination after exposure for 218 days, to a temperature
of 100°, while after 65 days, at 300°, they found in three tests
that 9°5, 0-4, and 1°3 per cent of the gas had united. They
obtained at 350° and 50 hours, 1°6 per cent, and after 120
hours, 1:9, 16°4, 0°5, 0°7 per cent. Krause and Meyer** found
_ that 28 to 100 per cent of detonating gas combined when
heated to 518° for 2 hours, in sealed bulbs. On the other
hand, Askenasy and Meyer++ observed that but little water was
formed where the gas passed slowly through a vessel at 606°.
Freyer and Meyert{{ determined the temperatures at which mix-
tures of gases exploded when slowly passing through a hot
vessel, and also when quickly heated in sealed bulbs. Explo-
sions occurred at the higher temperatures given in the follow-
ing table, but not at the lower:

With free In sealed
current, bulbs.

Hydrogen ) 650—730° 518-606°
Methane | with 650—730° 606-650°
Ethane | oxygen 606-650° 518-606°
Ethylane r required 606-650° 518-606°
Carbonic oxide | for complete 650-730° 650-730°
Hydrogen sulphide | combustion 315-320° 250-270°
Hydrogen and Chlorine 430-440° 240-270°

* Jour. Chem. Soc., xlv, 587. + This Journal, iv. 51.

t J. Chem. Soce., xlix, 384. ~ § Monatsheft f. Chem., xviii, 6, xix, 299.

|| This Journal, iv, 51, 1897. 4] Ber. d. deutsch. Gesell., xxviii, 2804.

** Liebig’s Ann., eclxiv. 85 tt Liebig’s Ann., cclxix, 49.

tt Zeitschr. fiir phys. Chem., xi, 28.

Explosive Gases and Gaseous Mixtures. 329

‘They also found that detonating gas suddenly heated in un-

_ sealed bulbs exploded at the same temperature as in sealed.

Before discussing chemical changes let us recall the condi-
_ tionof matter in gases—the molecules are moving very rapidly,
- some faster, some slower, an individual molecule moving in a
straight line for a very short distance, until it strikes another.
These collisions of one molecule with others are very frequent.
The atoms composing a molecule are also moving in respect to
each other, this motion constituting a portion of the internal
energy of the molecules, which is continually diminishing
through radiation. When the gas is heated the mean velocity
ot the molec#les is increased, and at the same time the atomic
motion and internal energy are also increased. Ata sufficiently
high temperature some of the collisions are so violent as to dis-
rupt the molecules and their atoms part company.

When a mixture of gases is heated, chemical combination
results from the encounters of different kinds of molecules
having motion of translation, or of the atoms, or of both. At
any given temperature the mean velocity of molecules may be
valenlated, but we have no data for determining at what
greater velocity the molecular collisions are followed by chem-
ical union. Polyatomic molecules are dissociated by heat, but
since there is no measurable dissociation ot hydrogen or oxygen
at 1600° it does not appear that dissociation by heat plays
much part in the change. Whatever the chemical changes
caused by heat in gases, whether combination or dissociation,
the condition necessary for these changes results from mole-
cular impacts. If, for example, detonating gas be heated, a
part of the molecules acquire a velocity and “internal energy
adequate for combination. Some of these encounter each
other and combine to form water, the free atoms of oxygen
uniting with each other or with hydrogen. The nascent water
molecules have in general a high velocity, and collide with
those of oxygen and hydrogen ; but it is not probable that a
single impact ‘of a new water molecule with one of hydrogen
or oxygen imparts to the latter energy adequate for combina-
tion. If it were so, the change would proceed with accelerat-
ing speed at a temperature of 300° to 500°. This view, that a
nascent water molecule does not necessarily cause hydrogen
and oxygen to unite, accords with that expressed by the writer,*
that the heat of combination caused by the glow discharge of
electricity does not produce further combination.

Why is it that chemical union proceeds slowly at 500°, and
why does not the energy resulting from combination cause
further combination, and thus raise ‘the temperature of the sys-

* This Journal, vi, 218, 1898.

3830 = Miater— Partial non-explosive Combination of

tem to the point of explosion? Let us suppose the mixed
gases be heated from 500° to 501°. This increase in tempera-
ture will not cause them to explode. Assume further, that the
heat of combustion during one minute is sufficient to raise the
temperature of the gases 1°, provided no heat is lost by radia-
tion. Experiments have shown that this rise in temperature
will not cause explosion. But the gases heated by their own
slow combustion continually lose heat by radiation, and hence
are only slightly hotter than the surrounding medium. Were
it possible to prevent loss of heat, the temperature would rise
and the change proceed with accelerating speed.

The observations of Freyer and Meyer (loc. cit.), that explo-
sive mixtures gradually heated have a higher ignition point
than when suddenly heated, are difficult to explain, and further
experiments may be useful. Their results also indicate that
pressure is without influence, while A. Mitscherlich* found
that at pressures less than one atmosphere the ignition point
falls with the pressure. As pressure in so many cases influ-
ences explosion, it is remarkable in Freyer and Meyer’s experi-
ments that detonating gas in an open vessel, when quickly
heated, exploded at the same temperature as it did under a
pressure of two atmospheres or more. Mitscherlich’s results
also are the opposite of what we might expect.

Let us first consider the phenomenon of explosion of deto-
nating gas at a low pressure when the combination is not
complete. An electric spark or hot wire ignites the gas, the
water molecules about the source of heat collide with the
neighboring ones of hydrogen and oxygen, molecules of these
elements being shaken up by many impacts and acquiring —
energy adequate for combination, and thus uniting and adding
energy to the system. In this way the change propagates
itself. A part of the molecules of hydrogen and oxygen have
not combined. Why? When a denser gas is ignited, the
combination is complete, and as the temperature of burning
hydrogen is sufficient to dissociate water, we must assume that
some of the water molecules are broken up, and as the tem-
perature falls they reform, all of the hydrogen and oxygen
uniting before the temperature is too low for combination. In
the case of the rarer gas, the impacts to which a molecule of
hydrogen or oxygen is subject from the nascent water mole-
cules are less frequent than in the denser gas, and hence it is
not as likely to acquire the energy needed for chemical union.
Moreover, molecules possessing velocities adequate for combi-
nation encounter each other less frequently in the rare gas ;
hence the chemical change will proceed more slowly, and there
will be more time for the system to lose heat by radiation. A

* Ber. d. deutsch. Chem. Gesell., xxvi, 399.

Explosive Gases and Gaseous Mixtures. 331

molecule with internal energy adequate for combination loses
this energy if its free path is relatively long; that is, given
sufficient time though very brief, the molecule assumes a con-
dition unfavorable to chemical union.

Let us now see if other phenomena are explained by this
hypothesis, considering first the case of an electric spark in
detonating gas at a pressure too low for explosion. Here we
have an enormous number of molecules combining in the
luminous path of the spark. These radiate energy, part being
electrical in its origin, i. e., a portion which was derived from
the external work done upon the gas, and part resulting from
chemical union. These encounter neighboring molecules of
hydrogen and oxygen, imparting to some of them energy ade-
quate for combination. Of these last a part combine, and we
have now to consider a portion of gas consisting of water
molecules, as well as those of hydrogen and oxygen. Some of
the last named possess energy adequate for chemical union, but
their encounters are too infrequent to restore by the heat of
combination the energy radiated, hence the change proceeds
with diminishing speed and finally ceases to propagate itself.

Feeble sparks do not explode detonating gas at a pressure of
half an atmosphere or even more, but they cause slow combi-
nation. In this case it is not necessary to consider the possi-
bility of the hydrogen carrying the electricity, since water
molecules are formed, and we need only take into account a
system made up of a relatively small number of nascent mole-
enles of water, molecules of hydrogen and oxygen having
energy adequate for combination, together with a relatively
large number of molecules of these gases not having such
energy. Under these conditions the impacts resulting in
chemical union are too infrequent to maintain by the heat of
their combustion the energy lost by radiation, and conse-
quently the change does not propagate itself and cause an
explosion. |

Carbonic oxide and oxygen combine slowly at pressures less
than an atmosphere when subjected to feeble sparks. Here
too the impacts of molecules with a velocity adequate for
chemical union are infrequent, as is evident from the fact that
the change does not propagate itself.

In the case of cyanogen and oxygen either the feeble spark
causes no chemical union, which is improbable, or as in the
previous instances the change does not proceed throughout the
gas because of the infrequency of impacts of high velocity
molecules. |

The study of the phenomena of explosives involves the
thermal effect, but for the present purpose the heat evolved
by the explosion or decomposition of equal volumes will suffice.

332 Mixter—Partial non-explosive Combination of

Two grams and 2 volumes of hydrogen burn with 1 volume of
oxygen, with a thermal result of 58,700 calories, the product
gaseous; and 2 volumes of the mixture, therefore, give 39,134 c.
Likewise, 2 volumes of carbonic oxide and oxygen give 45,312
e.; 1 volume of hydrogen and 1 of chlorine give 22,000 e. ;
and 24 grams of solid carbon and 2 of hydrogen unite with
an absorption of 48,170 ¢. to form 2 volumes of acetylene,
which evolves this amount of heat when resolved into its
elements. In the same way, we have for cyanogen 683,700 ¢.,
for nitrous oxide 17,470 c., and for nitric oxide 21,575 ¢. set
free when these compounds are decomposed into their elements.
The following table gives the resultant heat of the combina-
tion of equal volumes of three mixtures, and of the decompo-
sition of four endothermic gases:

Resultant heat, Nature of product.
Hydrogen and oxygen ._-... 39134 Gaseous
Hydrogen and chlorine ------ 22000 | ps
Carbonic oxide and oxygen-_- 45312 bee
Acetylene 22 ket ee . 48170 12/18 solid
Cvanoceny ies: Jee. ee See 65700 120 eee
Nitrous oxide! 4a eaeees - 17470 Gaseous
INMETIC OR TCO' oly SRE ii F

The molecular heats at constant volume of H,, O,, N,, CO,
NO, and HCl are 4:2 to 5, and of N,O, H,O, and CO,, about
one-half larger. Those of O,H, and C,N, do not appear to
have been determined, but they are probably not greater than
10. Evidently the molecular heats have little influence in
determining the temperature at which explosion occurs, and
the resultant heat does not account for the fact that the igni-
tion temperature of chlorine and hydrogen is much lower than
that of hydrogen and oxygen, or of carbonic oxide and oxygen.

Aeetylene requires a pressure 20 times greater than detonat-
ing gas for explosion, while the heat of decomposition of the
former exceeds that of chemical union of the latter. At
temperatures not high enough to cause acetylene to separate
completely into its elements or cause detonating gas to explode, |
the two gases exhibit similar deportment in this respect,
namely, that the change is one of slow combination. Acety-
lene polymerizes to form C,H,, C,H,, and other condensation
products, and detonating gas forms water. But acetylene,
when sparked under sufficient pressure, decomposes completely
into carbon and hydrogen. The impacts of acetylene mole-
cules with a certain velocity causes them to combine with each
other, but when they have a higher velocity complete decom-
position results. The change started by a spark in dense acety-

io y

— i
ears 3 be Bi 5

tee” 7 a Soa oe a a ee

Cw, -
SEE?

r 7

a
“en
ie! -
>
ie,
\. a
4
te
‘

i
<

Explosive Gases and Gaseous Mixtures. 333

lene propagates itself throughout the gas. About the path of
the spark there are sufficiently numerous encounters of mole-

cules with a velocity adequate to cause decomposition, and the

heat resulting is not lost by radiation before other molecules
of acetylene are decomposed. Thus the change continues
until only solid carbon and stable molecules of hydrogen
remain. A proposed method for the commercial production
of lampblack depends upon this fact. When, however, a

spark passes through the gas at common pressure, the change
occurs only in the path of the spark. Here the impacts of

molecules with a velocity adequate to cause decomposition are
more infrequent than in the denser gas—too infrequent, as the
result shows, to propagate the change. The heat about the
path of the spark is dissipated by radiation before it decom-
poses neighboring molecules. The infrequency of impacts
allows sufficient time for radiation, and for the molecules to
assume a stable condition. It is also probable that some more
stable molecules, such as C,H,, are formed. If acetylene issu-
ing from a tube is lighted, the decomposition does not extend
into the tube for the same reason that it does not extend
through the gas when sparked at ordinary pressure. .

The deportment of acetylene under high pressure, when sub-

jected to small sparks (p. 324) is similar to that of detonating

gas.* Condensation products are formed in one case and
water in the other. The minute spark imparts to the mole-
cules of acetylene energy adequate to effect combination, and
as the spark is visible we assume that the portion of the gas
which is glowing has a temperature at which acetylene decom-
poses when larger quantities are heated. Thus the infrequency

_cf impacts of molecules of high velocity is the real reason why

feeble sparks do not start an explosion in dense acetylene.

The non-explosion. of cyanogen (p. 325) under high pres-
sure when sparked, although itis more endothermic than acety-
lene, is doubtless connected with the fact that it requires a
higher temperature for its decomposition than the latter; or it
may be that the union of CN with CN is exothermic, while
the radical itself is endothermic. Nitrous oxide and nitric
oxide are stable at high temperatures and these also do not
decompose with explosive violence when sparked under pres-
sure. It is probable that all three of these endothermic gases
would explode if sparked under pressure and at a temperature
approaching that at which they dissociate.

* This Journal, iv, 51.

334 Mixter—Partial non-explosive Combination, ete.

_ Recapitulation.

Detonating gas, a mixture of carbonic oxide and oxygen, one
of cyanogen and oxygen, and other explosive mixtures of gases
do not explode below certain pressures when sparked: and the
decomposition of acetylene does not propagate itself at pres-
sures less than two atmospheres. In the path of-the spark
chemical changes occur which do not extend throughout the
gases, and the same is true of a weak spark in the gases men- |
tioned when under a pressure at which a strong spark explodes
them. It appears for reasons already stated that explosions do
not occur in the instances given only because of the infre-
quency of impacts of molecules having a velocity or internal
energy adequate for chemical union. In the rare gas the
impacts are less frequent than in a dense gas, the mean free ©
path is longer, and there is more time for a molecule with
energy adequate for combination to lose this energy by radia-
tion and to attain a condition unfavorable to chemical union.
Some of the molecules combine, but the heat of their union is
not sufficient to restore the energy lost by radiation and the
change is therefore not self-propagating. The same explana-
tion holds good for the phenomena of feeble sparks in a dense
explosive gas where in the path of a spark there are relatively .
few molecules with energy adequate for combination, and
these collide with each other less frequently than with the
molecules not having such energy. In a dense gas a given
quantity and a degree of heat, that is,-suificient frequency of
molecular impacts, is requisite to secure spontaneous extension
of the change. |

Williams— Occurrence of Paleotrochis in Mexico. 335 —

Art. XXXVI.—On the Occurrence of Paleotrochis in Vol-
canie Rocks in Mexico ; by HENRY S. WILLIAMS.

Introductory Note.—This article was written, Nov. 22,
1897, at the time of the study of the specimens sent by Dr.
Dugeés, but for various reasons was not then published. A
discussion in the Geological Society last December led the
writer to refer to these specimens as proof of the non-organic
nature of Paleotrochis. Later the Mexican specimens, together
with the manuscript copy of this paper, were sent to Mr.
Walcott and then to Mr. Diller in confirmation of the inor-
ganic origin, and association with igneous rocks, of Paleotrochis.
It is thought best to publish these facts for what they are
worth, in connection with Mr. Diller’s thorough study of the
subject as presented in the following paper.

SOME specimens sent to Yale College for identification, by
Dr. Alfred Duges of Guanajuato, Mexico, throw some light
upon the nature of the peculiar forms described by Emmons
under the name of Paleotrochis. The specimens come from
what appears to be an old eroded volcanic cone made up of
“Porphyre altéré,”’ ‘“ Peckstine rouge” and volcanic tuffs,
situated northeast of Guanajuato in the Santa Rosa Mountain
range. Among the specimens are several fragments of much
altered fragmental as well as massive volcanic rocks. One
specimen is filled with amygdules. |

The majority of the specimens are composed of milky quartz
of the shape called Paleotrochis by Emmons, with variations
in form which are particularly suggestive. These suggestive
variations are the imperfect form, one-sided, or as if made up
of several of the conical halves of typical paleotrochis joined
together at their bases, some of the cones being very broad
and nearly flat, but each having the appearance of accretion
from the small end which in the lower cones is near the center.
When thus nearly flat the lines of concentric accretion are
evident, while the cones with less angle of slope show the
radiating lines more conspicuously. The suggestion furnished
by their study is that they were formed in a manner similar to
that of the formation of ice columns under a loose soil, which
at about freezing temperature sometimes grow to several inches
in length. The lengthening of the column of ice, in the latter
case, appears to be due to additions of ice forming as soon as
the cool atmosphere is reached, while the temperature of the
soil is still enough below freezing to permit the water to be
drawn up by capillary flow to the surface as the freezing
removes it from the surface immediately below.

336 Wilkams—Occurrence of Paleotrochis in Mewico.

In the case of the Paleotrochis the application of the sug-
gestion would be in this wise: If a porous rock, had arising
through it from below, superheated waters carrying solution
of silica, a point might be reached when suddenly the loose-
ness of the rock would permit of rapid radiation upwards of
heat, while below the temperature of the heat would be high
and remain nearly constant for a long period. . Such conditions
might occur along the plane of a considerable exit of the rising
waters, the rocks below being saturated with the heated waters
and above the rock being more or less open by the absence of
water. J conceive that, at the junction, a condition might
occur in which silica would begin to deposit on the interior of
a cavity, but the supply of silica, or of silicated waters, would
be furnished only as it was abstracted by solidification. Thus
at the points where solidification began the cavity would be
increased by the crystallization, and into this cavity more sili-
cated waters would arise; thus the accretion-surface would
increase with the lengthening of the cone by the separation of
the apices of the cones; consolidation taking place at the
point of meeting of the solidified silica with the water, and
the water reaching that point slowly. 3

If some such explanation be the fact, the association of these
peculiar forms with gold is explained. Also that such condi-
tions should occur in volcanic regions and particularly in the
midst of voleanic tuffs, as in the Mexican case, is as would be
expected.

Nov. 22, 1897.

J. S§. Diller—Origin of Paleotrochis. 337

Arr. XXXVII.—Origin of Paleotrochis ; by J. 8. Druze.

Pror. EBENEZER EmMons,* while State Geologist of North
_ Carolina, discovered among the so-called Taconic rocks of
Montgomery County, in that State, a number of more or less
regularly striated bi-conical forms to which he gave the names
Paleotrochis major and minor, and regarded them as siliceous
corals as well as the oldest representatives of animal life
upon the globe. According to Emmons Paleotrochis varies
' in size up to two inches in diameter, aud occur with
many almond-shaped concretions, often within concretions, ina
series of beds over 1,000 feet in thickness interstratified with
beds of granular quartz, conglomerate and quartzite. Both
species of Paleotrochis have the form of “a flattish double cone .
applied base to base” with the surfaces grooved somewhat
irregularly from near the apex to the basal edge. The smaller
form, P. minor, has the “ apex of the inferior side excavated or
provided with a small roundish cavity” and the other apex —
“supplied with a small rounded knob, from the base of which
the radiated grooves begin.” The larger form, P. major,
“differs from the foregoing (P. minor) in the absence of the
_ roundish apical depression of the lower side and the knob of
_ the opposite side.” re
Prof. Emmons regarded Paleotrochis not only as originally
siliceous but also gemmiferous, thus accounting for knobs as
well as irregular adhering groups, and it is important to note
that he reports “these fossils also occur in a variety of quartz
or quartzite which [ have described as a buhrstone, and which
is often porphyrized.” |

Prof. James Hall,+ after an examination of many specimens,
regarded the Paleotrochis of Emmons as nothing but concre-
tions in quartz rock. Prof. O. C. Marsht examined the forms
microscopically and found them composed of fine-grained
quartz, but no trace of organic structure could be detected.
While maintaining its inorganic nature he regarded it as diffi-
cult to explain, and considered it as having some analogy with
cone-in-cone, which he thinks is probably due to the action of
pressure on concretionary structure when forming.

The most extensive paper on this perplexing form is that of
Mr. C. H. White,$ who strongly advocates the organic nature

* Geological Report of the Midland Counties of North Carolina, 1856, page 62,
also this Journal, II, vol. xxii, page 390, and vol. xxiv, page 151.

+ This Journal, II, vol. xxiii, page 278.

¢ This Journal, II, vol. xlv, page 218.
- § Journ. of the Elisha Mitchell Scientific Soc., Chapel Hill, N. C., Part 2d, July
to December, 1894, pages 50-66.

338 J. S. Diller— Origin of Paleotrochis.

of Paleotrochis. The specimens he examined were those
obtained by Prof. Emmons, as well as a number collected by
Prof. J. A. Holmes, the present State Geologist of North Caro-
lina. Mr. White describes in detail not only the peculiarities
of the weathered surface of the rock but also the features
exhibited upon a fresh fracture, and called attention for the
first time to the radial fibrous mineral which he regarded as
impure chalcedony. According to Mr. White, the fossil forms
are enveloped in chalcedony and the small concretions are made
of the same material.

In 1887, Prof. J. A. Holmes* visited the Sam Christian

Gold Mine of Montgomery County, N. C., and studied the
Paleotrochis-bearing rock in the field. Although he had not
then seen any of the acid volcanic rocks from New England,
described by Dr. M. E. Wadsworth, or from the South
Mountain region of Maryland and Pennsylvania, subsequently
described by Dr. G. H. Williams and Miss Florence Bascom,
he was of the opinion that the rocks in the neighborhood of
the Sam Christian Gold Mine were of eruptive origin. Later
observations have convinced him of the correctness of this
view. The same opinion is entertained by Messrs. H. B. C.
Nitze and George B. Hanna,+ who consider the Paleotrochis- _
bearing rocks at the Sam Christian and Moratock Mines as _
ancient acid volcanics, and state that “it appears highly prob- —
able that at least some of these siliceous pebbly concretions are
spherulites.” Unfortunately in the preparation of their report
time did not permit the authors to study thin sections.
The specimens which, at the request of Mr. ©. D. Walcott,
the Director of the U. S. Geological Survey, the present
writer has had an opportunity to study, consist of a small col-
lection{t from Mexico sent by Prof. H. 8. Williams, besides
three fragments about nine inches in diameter sent by Prof. J.
A. Holmes, who collected them in 1887 at the Sam Christian
Gold Mine, North Carolina, and from the same place several
dozen of the original specimens of Paleotrochis major and
minor collected by Prof. Emmons. Specimens of the rock
and isolated fossils, excepting those from Mexico, have been
cut and polished and thin sections prepared for microscopical
study.

The rock from North Carolina which contains Paleotrochis
is full of nodules of various shapes and sizes, ranging from
that of a pin’s head to nearly two inches in diameter. These
are the supposed concretions and fossils. Upon a fresh fracture
the rock appears to be composed chiefly of quartz, but when

* Letter to the author Feb. 6, 1899.

+ North Carolina Geological Survey, Bul. No. 3, pages 37 and 39,
t See Prof. William’s article, this volume, page 335.

*

J. S. Diller— Origin of Paleotrochis. 339

weathered most of the nodules become white as if kaolinized,

_ while the other nodules and the matrix remain quartzose in

appearance. ‘The nodules form at least two-thirds of the mass
of the rock and are arranged with their longer diameters paral-

lel, rendering the rock rather easily split in one direction.

With a lens, it may be seen that the small kaolinized nodules

exposed in section upon the surface of a hand specimen have a

radial fibrous structure. The same structure may be seen in

some of the larger ones, and in addition to this feature some

of the nodules possess a more or less distinct concentric shell- \
like structure. These structures are usually best displayed

_ upon or close to a weathered surface. Portions of the nodules

or spaces between them are in a few eases cellular, and the
walls of the openings are rarely lined with minute crystals.
The supposed fossil forms usually appear conical or discoidal
upon a weathered surface. They often show a small cup in
the apex and are surrounded by a narrow depression from
which the radial fibrous envelope pointed out by Mr. White

has been removed by weathering.

A careful comparative study of the nodules in the hand
specimens tends to convince one that however different in form
and size the supposed fossils and concretions may appear, all
belong to one series and have essentially the same origin.

A microscopical study of thin sections of the rock reveals
the fact that the nodules are spherulites, a common feature of
acid igneous rocks. They are composed in most cases chiefly
of fibrous feldspar with quartz or tridymite. As seen in the
thin section of the Paleotrochis-bearing rock, the fibers are
grouped radially with more or less irregularity in tufts, sheaves,
sectors, hemispheres or spheres. When they form a complete
sphere, which is rarely the case, they are most coarsely fibrous
or granophytic at the center and usually show between cross
nicols an indefinite black cross. Occasionally also the concen-
tric structure is well marked. The rays are too minute to
permit of an accurate determination of their mineral composi-
tion by optical, methods, but microchemical tests with hydro-
fluosilicic acid yield the small cubic erystals, characteristic of
potassium fluosilicate as well as the hexagonal prisms of sodium
fluosilicate. Judging from the greater abundance of prisms
than cubes the fibrous feldspar is richer in sodium than potas-

sium. ‘That feldspar, instead of chalcedony, is the most prom-

inent constituent of the spherulites, is fully borne out also by
its kaolinizing under the influence of the weather.

The spherulites are embedded in a matrix composed chiefly
of granular quartz. The grains are occasionally so large that
the uniaxial positive character can be readily determined.
Untwinned feldspar in small grains may be present in con-

Am. JouR So1.—FourtH Series, VoL. VII, No. 41.—May, 1899.
23

340 J. 8. Diller—Origin of Paleotrochis.

siderable amount and yet be easily overlooked. The quartzose
character of the weathered matrix, however, shows that at
least where most coarsely granular there cannot be much if
any feldspar present in it. In places the matrix contains
numerous minute parallel scales of what appears to be sericite.
Associated with the most coarsely crystalline areas are a few
scales of brown biotite and occasionally considerable green
biotite, which in places is so abundant as to make quite promi-
nent dark green spots. Both matrix and spherulites are trav-
ersed in a few cases by small veins of granular quartz, showing
that there is a considerable amount of secondary quartz present.
Both spherulites and matrix are rendered slightly micropor-
phyritic by containing occasional crystals of plagioclase,
feldspar and quartz. ‘The plagioclase, which, on account of
its small angle of symmetric extinction, must be an acid one,

in some cases forms the center from which the spherulitic —

fibres radiate.

An isolated specimen of Paleotrochis was cut through the 4

apices and found to be composed of granular quartz. The
quartz was fine-grained upon the outside where the grains were
set with their longest axes perpendicular to the adjoining sur-
face. The middle portion contained an irregular iron-stained
cavity possibly due to the disappearance of some iron-bearing
mineral. Several of the half embedded forms of Paleotrochis
were cut in a hand specimen to discover its relations to the
enclosing rock, and in each case it formed the interior portion
of aspherulite. Most of them contained a dark green patch.
The exposed conical surface of one was well striated and there
was an irregular depression in the apex. The form was com-
posed chiefly of granular quartz with a yellowish brown to
dark green, strongly pleochroie ‘biotite. Near the center is a

small spherulite which is not only bordered by finer-grained —

quartz but is cut by a small vein of it, showing that the depo-
sition of the quartz is subsequent to the development of the
spherulite. ‘The embedded portion of Paleotrochis is bordered
by spherulitie fibers which run approximately parallel to the
slope of the conical surface, and it is evident that the casts of
these fibers produce the irregular striz or grooves upon the
surface of the supposed fossil. The embedded portion termi-
nated with an irregularly-pointed apex below. The whole
form is fine-grained near the border and sends small veins into
the adjoining spherulitic shell. These veins areso small as not
to be visible upon a polished surface of a hand specimen even
with the aid of a pocket lens, but come out distinctly in the
thin section. The spherulitic shell by which Paleotrochis is
enveloped is composed of fibers belonging to a number of
centers or lines and yet combined they appear to form one

J. 8. Diller—Origin of Paleotrochis. 341

nodule. The biconical form of Paleotrochis suggests that it
originated as two spherulite sectors of which the apices were
the centers from which the fibers radiated. This would seem
to be the simplest way to account for the most regular as well
as many of the irregular forms, but of the specimens examined
I have not been able to find one that certainly originated in
that way.

A number of the fossil forms with a well-marked cup in the |
exposed apex turned ont to be flat hemispherical or thin Jen-
ticular in section, and are composed wholly of spherulitic fibers.

Although admitting much irregularity especially on account
of the supposed gemmiferous character of Paleotrochis, the
ones which have been considered the most characteristic of the
fossil are the distinctly biconical forms. These so far as seen
are chiefly granular quartz with more or less green biotite.

It is important to note also that the dark groups of green
biotite occur in the interior of very irregular nodules which
have ro suggestion in them of Paleotrochis. Irregular filat-
tened lenticular masses of granular quartz with green biotite
occur within the spherulites as well as about them. The
green mica is found only in the most coarsely granular groups
of quartz.

The following chemical analysis, made by W. I. Hillebrand,
shows that the rock has the composition of a rhyolite and
accords closely with the results of the microscopical study.

Analysis of the Paleotrochis-bearing rock of Sam Christian
Mine.*

ern aes! TOD
bee sere ie Le “Pl
1 Eas ee 11°41 with a very little P,O,
oS: Meee Cae "20
gt oS ae 70
TS OS Sa eee none
CaO 9]
SrO 0 a aaa aes
SS Ss Sea eee 05
CTS So eee a very little
[ee 3°52
10 2 ee 3°46
PLO below 105° -...-< "18
PasasvOve. < . "OF Gueminion )
100°02

Recognizing the Paleotrochis rock as an acid voleanic full of
spherulites, it is easy to understand the great variation in the
form of the nodules. Such rocks are in many places distinctly
banded and were long considered siliceous sediments, but by

* No other constituents looked for.

342 J. 8S. Diller— Origin of Paleotrochis.

the investigations of Wadsworth, Williams, Bascom and others
it has been definitely settled that they are all acid voleanies.
These rocks in North Carolina are regarded by Mr. Holmes
as pre-Cambrian and since their eruption may have undergone
great changes like those of the South Mountain described by
Miss Bascom. Some of the supposed fossils are certainly
spherulites, and all of them may have been originally. Some
broken forms show motion in the mass after the spherulites |
were developed. That Paleotrochis where most perfectly de-
veloped and composed of granular quartz is the result of
deposition, after the spherulitic growths about it and within it
had developed, there can be no question, but whether this depo-
sition followed soon after that of the spherulites in the course
of solidification or took place in hollow spherulites (lithophysee),
or resulted perhaps long subsequently at the time of rock altera-
tions, is not so clear. All this and much more will doubtless be ©
cleared up by the members of the Geological Survey of North
Carolina, who were the first to correctly ‘identify the rock and
the character of the supposed fossil.

None of the Mexican specimens received from Prof. Wil-
liams were cut for microscopical examination. Some of them
were clearly of igneous origin, and contained amygdules. The
Paleotrochis-like forms with radial markings appeared to be
composed of secondary quartz and probably originated as those
of North Carolina.

About a year ago bi-conical forms like Paleotrochis were pre-
sented by Mr. Kochibi, Director of the Geological Survey of _
Japan, to the U.S. Geological Survey. These specimens are —
now in the National Museum, and are much more regular in
form, size,,and general appearance than the Paleotrochis of
North Carolina. They are of a pale pink color with regular
bi-conical, striated forms, and in some cases have shallow pits
in one of the apices. They are known in Japan as “Soroban
ishi” or abacus stones. One of these specimens contains a
small fragment of the rock from which these curious specimens
were obtained, and it appears to be spherulitic. According to
Mr. Willis, who obtained the information directly from Mr.
. Koehibi, “ these stones are found only in rhyolitic tuffs. They
not infrequently occur much larger than these specimens, pos-
sibly up to two inches in diameter or more, and are more fre-
quently associated in groups of two or three overlapping or
coalescing. ‘They are generally white, the rosy tint of these
specimens being a rare characteristic.” A thin section of one
of these ‘ abacus-stones ” shows it to be an agate of which the
outer layers are pink and the inner white. There can be no ?-
doubt in this case that the form resulted from the filling of the
cavity long after the solidification of the igneous material. ——~

_ U.S. Geol. Survey, Washington, D. C., March 3, 1899.

O. A. Derby—Association of Argullaceous Rock, etc. 3438

Art. XXXVIII.—On the Association of Argillaceous Rocks
with Quartz Veins in the Region of Diamantina, Brazit ;
by ORVILLE A. DERBY.

IN a recent excursion in the region about Diamantina, Minas —
Geraes, Brazil, the frequent association of the decomposition

__ produets of argillaceous rocks with the quartz veins with which

the underlying metamorphic series of the district is threaded
attracted attention. The phenomenon was the more striking
because of its absence in a long stretch of road over a newer
series in which quartz veins are equally if not more abundant.

The most frequent appearance is that of a selvage to the
veins and of intercallations in the mass of the quartz of a red-
dish foliated clay apparently resulting from the decomposition
of a micaceous rock. Less prominent in the surface exposures
but quite frequent in the artificial openings that have exposed
the inner portions of the veins, are nests and stringers of a
structureless white clay (kaolin) that may be either farinaceous
or indurated (lithomarge). The red clay was observed in a
sufficient number of instances to warrant the assertion that it
is a general, if not a universal, feature of the quartz veins of
the region and that when it occurs it is constant throughout
the entire extension, both lateral and vertical, of the vein.
The distribution of the white clay, on the other hand, is patchy
and considerable portions of a vein in which it is known to
occur may be free from it.

A magnificent exposure of a vein of this character is afforded
by the eastern wall of the great cutting of the diamond mines

_. of Sao Joao da Chapada, of which an excellent view, repro-

duced from a photograph, is given in Boutan’s monograph on
the diamond (Encyclopédie Chemique de Fremy), and in a
recent paper by the present writer in the Journal of Geology
(vol. vi, p. 121). Contrary to the description previously given,
based on a rapid examination made under unfavorable circum-
stances and with a limited experience in the identification of
decomposed material, this cutting is made along the junction
of a conglomeritic diamond-bearing quartzite, and an older
fine-grained quartzite (itacolumite), both being profoundly de-
composed and in many places indistinguishable except on the
closest scrutiny. The eastern wall, which on account of its
inaccessibility in the wet season, when it is defended by a
frontage of quicksand, and of its superficial staining from the
highly-colored overlying ‘soil-cap has been described as a com-
plex of various schists, proves to be a uniform mass of itacol-
umite with only a single argillaceous member that presents in
a striking manner the above-mentioned association with vein

344 O. A. Derby—Assocration of Argillaceous

quartz. Throughout the entire length of the cutting, a dis-
tance of over 500 meters, runs a sharply-defined layer from 1
to 2 meters thick of argillaceous material with quartz which
has received from the miners the name of Guza (Guide),
because, as they state, diamonds were to be looked for below
‘(or in front) and not above (or behind) the outerop of this
layer.

This layer consists principally of a mass of dark red foliated
clay which is evidently a decomposed schist stained with oxide
of iron. In several places that afford access to its interior, the
central portion is seen to be occupied by an irregular but
apparently continuous mass of vein quartz, that sometimes
swells out so as to occupy half or more of the thickness of the
layer, being in other places reduced to a thickness of a few
centimeters only. Associated with the quartz is a pure white
granular kaolin which in places is intimately mixed with small
angular granules of quartz in a way that suggests a decom-
posed granite, but in general forms pockets or a selvage in or
around the more massive portions of the vein. Outside of the
kaolin, when this forms the marginal portion of the vein, or
inclosed as angular fragments in its mass, is a minutely banded
reddish, or yellowish, micaceous clay that appears to differ
from the red clay composing the greater part of the mass of
the Guia only in the absence of the intense staining with
oxide of iron. The relations of the white kaolin with the
banded clay is well shown on the scraped surface of a block
that was bronght away, and represented about one-fourth the
natural size, in the accompanying figure, in which the dotted

portion represents the granular white kaclin and the shaded
portion the banded clay. The unshaded streaks in the latter
are also of white kaolin that differs from that of the vein in
the less distinct granulation, but which in places, as at the lower
right hand side, appears to be flame-like apophyses from it.

Rocks with Quartz Veins in Brazil. 345

A very pure specimen of the granular kaolin kindly analyzed
by Dr. Hussak gave: SiO, 44°96 per cent, Al,O, 42°09 per
cent, H,O 13°25 per cent. This composition differs from that
of a typical kaolin derived from a feldspar only in a slight
excess of alumina which may perhaps be due to the precipita-
tion with it of rare elements that are known to occur in the
clay.
The quartz and kaolin are so blended that they must be
considered as constituting a single geological body, and this
ean hardly be else than a rock of original granitic, or rather
_ pegmatitic type. Moreover this body from the character of
its contact, as shown in the figure, must have been eruptive.
The character of. the accessory elements is in accord with this
view. Those that appear macroscopically are rare prisms of
red rutile and handsome tabular crystals and groups of specu-
lar iron. The heavy residue obtained by washing is in parts
very small, in other parts (where rutile abounds) very abundant,
and with the appearance of being wholly authigenetic. It con-
sists principally of rutile partly in rather large sagenitic groups
of octahedral form that are evidently pseudomorphs after some
mineral of the regular system (possibly titaniferous magnetite)
with extremely rare grains of anatase, magnetite, tourmaline and
xenotime. ‘The latter isin beautiful glassy prisms with short
pyramidal terminations of rather large size that show a ten-
dency towards crystals of macroscopic dimensions such as occur
rarely in the concentrates of the mine. The rutile and anatase
are evidently secondary, and of the presumably primary accesso-
ries, magnetite, tourmaline and xenotime, the latter is the most
significant as thus far it is only known 7m sitw, at least in
Brazil, as an extremely constant and characteristic accessory of
the ultra-acid (muscovite) granites and pegmatites. The ab-
sence (at least apparent, as in order to avoid admixtures only
smal] quantities could be washed) of its almost constant com-
panions, zircon and monazite, is noteworthy. In this connec-
tion it may be recalled that in a recent paper by Dr. Hussak,*
on an auriferous quartz vein in the same series near Ouro
Preto, these three characteristic granitic accessories were found
together, and from this circumstance and from the occurrence
of characteristic contact minerals the same conclusion was
arrived at, viz.: that the vein is an ultra-acid phase of a
granitic apophyse.

The colored clays of the Guia present the appearance of the
decompusition products of a micaceous schist. As already
stated, the distinct banding of white and colored portions, such
as is represented in the figure, in the parts next to the quartz-
kaolin vein, is probably due to minute interlaminated projec-

* Zeitschr. f. prakt. Geol., Oct., 1898.

346 O. A. Derby— Association of Argillaceous

tions from the latter, and in support of this view is the fact that
the white bands are distinctly kaolinitic (though not granulated
like the kaolin of the vein), while the colored ones are as dis-
tinctly micaceous. Aside, however, from these distinct white
bands, there is an obscure lamination visible even in the most
highly-stained portions. On a carefully scraped surface cut
transverse to the lamination a few scattered rounded eyes of
quartz are in some portions visible, which with the banded
structure are very suggestive of a stretched quartz porphyry.
_ These eyes occur both in the dark red clays and in the lighted
colored ones, but are much more abundant in the latter than
in the former. This fact taken in connection with the in-
creased staining of the clay towards the margins of the mass
and the apparent lack of a physical break between the lighter
and the darker colored clays may, perhaps, be taken as an
indication that the original rock was more acid in the central
and more basic (or at least richer in iron) in the marginal por-
tions of the mass.

On washing, both the dark and the light colored clays give,
after a considerable loss as slime, an abundant residue of white
micaceous aggregates (in the case of the red clay treatment
with acid is necessary) with the aspect of sericite with rare
granules of quartz that appear to come exclusively from the
eyes. On separation of the mica and quartz by washing or the
use of heavy liquids, we find an extremely abundant residue of
mixed rutile and anatase (sometimes the one, sometimes the
other predominating), which appear to come from original
ilmenite represented by abundant black grains in the red and
by rather rare, dirty aggregates of ill-formed anatase or rutile
in the lighter colored clay. With the titanium minerals occur
rarely minute prisms of tourmaline and prismatic fragments,
rarely perfect crystals, of monazite of the peculiar type to be
described farther on. All the elements of the residue are dis-
tinctly authigenetic and in this respect are in marked contrast
with those of the adjoining quartzite, which with the same
titanium minerals (with the exception of ilmenite) and tourma-
line in perfectly fresh crystals contains an abundance of well-
rolled zircons.

The Guia is thus a composite body that consisted originally
of a vein, or dike, of a quartz-feldspar rock, presumably
pegmatitic granite, injected into a micaceous schist that may
have been originally either a sedimentary shale or a sheared
eruptive. The appearance of the mass is exceedingly dike-like,
but if a dike, it must have been injected along a bedding
plane, or so nearly coincident with one as to simulate an inter-
callated layer. The characters above given of apparent por-
phyritic structure, absence of distinctly clastic elements, and

Rocks with Quartz Veins in Brazil. 347

the presence of monazite, all favor the view of an eruptive
origin and appear to be of more weight than those of schistose
structure and apparent interbedding with the quartzite which
are the most important that can be cited in favor of the oppo-
site view. This question will be discussed farther on.

For an attempt to reconstruct the original rock of the colored
portion of the Guia we have the positive indications of a schist
free or nearly free from quartz and rich in sericite (or a mica-
ceous mineral resembling it which for convenience will be
ealled by that name), with an abundance of iron and titanium
minerals, tourmaline and monazite, as characteristic primary
accessories and with an entire lack of distinetly clastic elements
which in a series with the degree of metamorphism of that
about Diamantina could only be rolled zireons or regenerated
tourmalines.* Tor the purpose of attempting such a recon-
struction, specimens were collected of such rocks as from their
general aspect might be suspected of representing the original
type from which the Guia clays might be derived. From the
numerous specimens examined from various localities, the com-
parative freedom from quartz and the abundance of sericite,
of iron and titanium minerals and of tourmaline prove to
be common characteristics of the phyllites of the region, which
can only be “cane aan as presumably original sedimentaries
or eruptives by the presence or absence of distinctly rolled
zircons (often of asize and abundance that seems extraordinary
in rocks presumed to represent fine-grained sediments) and of
monazites or zircons, or both, with the aspect of original
primary accessories.

The rock which from its position and aspect seemed to be
the most promising was a partially decomposed greenish schist
found in loose blocks at the foot of the western wall of the
cutting, and which, judging from appearances, was thought to
have fallen from a layer somewhat similar to the Guia and only
a few dozens of meters below it in the same geological series.
This, however, proves to be characterized by an abundance of
well-rolled zircons and of regenerated tourmalines and by an
entire absence of monazite.

Another rock from the immediate vicinity and presumably
from the deep rock cutting (mainly in hard itacolumite) for
the drainage canal of the Duro mine (seen in the background
of the above cited figure in the Journal of Geology), is a thor-

(a ee Ga es) a eR

a ee

* See Derby, On the accessory elements of itacolumite, ete., this Journal,

nite ay

frequently in the quartzites about Diamantina might have been derived from
original clastic phosphates. This view is confirmed by an examination of more
abundant material in which distinct but partially altered and corroded monazites
and xenotimes were found in the heavy residue of nodules rich in peer bande |
lazulite.

bo
2

%
ie
®

March, 1898. In this paper it was suggested that the lazulite that occurs quite’

348 0. A. Derby—Association of Argillaceous

oughly sound, bluish, imperfectly laminated phyllite of which
numerous blocks have been built into a wall at the mine-house.
Some blocks have in the joint planes handsome plates of specu-
lar iron like those of the Guia clay. The rock is for the most
part mottled with large yellowish blotches suggestive of
enclosed crystals or pebbles, and on a polished cross section
shows a minute banding with sheared eyes, corresponding to
the blotches, suggestive of a sheared porphyritic rock. The
most massive specimen shows evidence of shearing in two planes
nearly at right angles to each other, and, in this, part of the
original structure which, in the most of the specimens, has
been obliterated by the two systems of shearing can still be
observed. The appearance under the lens is that of an altered
diabasic rock sheared to the point of obliterating the rectangu-
lar outlines of the feldspars without completely destroying
them. The spots above referred to have the aspect of pheno-
erysts of feldspar that have lost something of their sharpness of
outline. Microscopic sections show a general sericitic mass
blackened in irregular bands and patches with a fine dust of
iron ores which on being dissolved with acid leave whitish
ageregates of a titanium mineral (anatase?). The heavy
residue consists of a great abundance of iron and titanium
minerals, the latter only rarely showing well-defined forms of
rutile and anatase. The only other elements recognizable are
rare grains of tourmaline and small prismatic fragments, rarely
perfect crystals, of monazite. Some of. the grains referred to
this mineral may possibly be zircon, but if so they are perfectly
fresh (not rolled), but indistinctly characterized. The original
rock was apparently a basic eruptive very rich in iron and titani-
um minerals (ilmenite?) and with rare phenocrysts of the more
acid element. So far as can be made out, this rock would on
decomposition give exactly the characters of the red clay of the
Guia. The principal difficulty in the interpretation of this
rock and of the red clay of the Guia as a sheared and meta-
morphosed basic eruptive is the presence of monazite, which
thus far is only known zm se¢éw in the acid eruptives or their
metamorphosed representatives. |

Two other rocks from the same region present certain char-
acteristics in common with the one above described and if, as
seems probable, they can be genetically connected with it, will
serve to establish a graduation into a more acid type on the
one hand (thus removing in part the difficulty above noted),
and on the other hand into a better-defined basic type. From
near the Serra do Gigante, several miles to the northward of
Sao Joao da Chapada on the road from Diamantina to Grao
Mogol, my friend and companion, Dr. Francisco de Paula
Oliveira, brought a loose block of a peculiar cyanite-phyllite

Rocks with Quartz Veins in Brazil. 349

full of enormous crystals of cyanite in an imperfectly laminated
groundmass of sericitic appearance. The rock is free from iron
ores, but extremely rich in minute aggregates of rutile and of
extremely minute prismatic needles that are referred to mona-
zite. The original rock was évidently a highly aluminous
(feldspathic) one, with an abundance of a titanium mineral
(ilmenite?) and of monazite, in which through shearing and the
development of secondary minerals, all traces of the original
structure has been obliterated. A very similar rock, but lack-
ing the monazite, was found by Dr. Hussak on the Ilha do Fogo
in the river Sao Francisco in front of the town of Joazeiro,
where it presents the aspect of a dike in gneiss.

Some 15 or 20 miles to the southward of Sao Joao da
Chapada, near the fall of the river Dattas close by the town of
Gurvéa, another imperfectly laminated phyllite was found out-
cropping in the road that strongly resembles that of Sao Joao
but differs principally in the absence of monazite, in the pre-
servation of the original rectangular outlines of the white
mineral transformed into sericite, and in the great quantity of
tourmaline that is evidently an element introduced in the
metamorphism of the rock. In addition to the fine iron dust
that on removal by acid leaves a dust of formless aggregates of
titanium, the rock contains magnetite in considerable-sized
grains and octahedral aggregates of an altered titanium mineral
that strongly resembles those of decomposed perofskite
described by Hussak from the diamond gravels of Agua Suja
(Neues Jahrb., 1894, ii, p. 297). Judging from the sharply
eut rectangular outlines of the areas of ‘sericite free from the
iron dust, the original rock must have been largely composed
of lath-shaped crystals like those of the plagioclase or melilite
of diabasic or basaltic rocks, and with this conclusion the great
abundance of the iron and titanium dust, that may be presumed
to come from original ilmenite, and of magnetite with an octa-
hedral titanium mineral, is in accord. The shearing has not
been sufficient to obliterate the form of the white element and
appears to have been anterior to the introduction of the tour-
maline that lines the joint planes and has invaded many of
the rectangular sericitic areas without passing beyond their
limits.

There can be no reasonable doubt but that the rock last
described is a metamorphosed and moderately-sheared basic
eruptive and when it is recalled that “mixed dikes” have
been described (see Zirkel, Petrographie, i, p. 784) in which a
granite-porphyry passes to.a type that has been described as
diabase or melaphyre, the hypothesis that the two rocks above
described and the decomposed material of the Guia may be
genetically related is not so extravagant as at first sight appears.

350 O. A. Derby— Association of Argillaceous

The most serious objection to bringing the Dattas rock into
line with the others is the absence of monazite, but in regard
to this it may be noted that this mineral (like the xenotime in
the white kaolin) is so rare that it failed to appear in some of
the tests made.*

The association of argillaceous material with quartz veins is
also a prominent feature in the diamond mines about the little
village of Sopa, about a dozen kilometers to the southward of
Sao Joao da Chapada. The diamond-bearing material is here
a decomposed conglomeritic quartzite entirely similar to that
now on view in the mines of Sao Joao.t This rests uncon-
formably on an older series of quartzites (itacolumites) with
intercallated schists that is also profoundly decomposed. This
lower series is abundantly threaded with small irregular quartz
veins that frequently show included pockets and “stringers of
indurated kaolin and almost invariably partings and selvages
of various colored schistose rocks or clays that closely resemble
those of the Guia above described.- The white kaolin which
is less abundant and constant than in the Guia, affords the
same residue of rutile and anatase with rare prismatic grains
of xenotime. In a washing from the kaolin of the Bamba
mine the residue, consisting almost exclusively of anatase with
some grains approaching a macroscopic size, was nearly 0°05%
of the quantity washed. In another from Misael’s mine the
residue is scanty, as in the greater part of the tests from the
Guia, and consists principally of rutile with rare grains of ana-
tase. In both the xenotime is in rare individuals, mostly in
fragments, of comparatively large size showing a tendency to
assume macroscopic proportions.

The schistose accompaniment of the veins at Sopa is, in the
few cases observed in which it is well preserved, a well lami-
nated, almost papyraceous, micaceous greenish schist that gives
on decomposition an ash-colored, or reddish, clay, not distin-
guishable, except by its residue, from the clays derived trom

* Owing to this circumstance, and to the difficulty of avoiding admixtures, the
tests of the material from the Guia in which the rare accessories proved to be
infrequent and inconstant, were many times repeated and with special precau-
tions. The identification based on form, optical properties and behavior with
heavy liquids was confirmed by microchemical tests for phosphorie acid and for
cerium and yttria.

+ The question of the diamond in these deposits will be discussed in another
place. The conclusion presented in former papers, that the gem occurs in vein
material. was based on the examination of a mass shown me in 1880 as a diamond-
bearing body, but which I am now convinced was a detached section of the Guia,
in which, according to the best obtainable information of the present day, no
diamonds have been found, and which presents no evidence of having been
worked. On the other hand, it is certain that one of the diamond-bearing bodies
not now visible was closely similar to it in appearance, though perhaps not in
original character and origin. The section of the Guia examined in 1880 was
much richer in tourmaline than the parts washed on the recent visit.

Rocks with Quartz Veins in Brazil. 351

the phyllites intercallated in the series in which the veins
occur. A large “horse” left standing in the center of Misael’s
mine, and decomposed to a yellow ocherous clay, shows to the
naked eye abundant pyrite (altered to limonite) and tourmaline,
and affords on washing a very slight residue of quartz, much
tourmaline in handsome prisms, and a comparative abundance
of well-rolled zircons with rare grains of rolled rutile and still

|_ rarer ones of sagenite that, like the tourmaline, appear to be authi-

CD TP

PS tz

hd ee

'?- =

eS te

genetic. An ash-colored clay poorly exposed, only afew meters
away and of very similar aspect, affords an entirely different
residue characterized by an extraordinary abundance of minute
sharp-angled prisms of monazite of peculiar type, with rare tour-
maline and still rarer octahedral xenotimes often intergrown with
zircon but altered to a yellowish mass that is difficultly distin-
guishable from the aggregates of titanium minerals so abundant
in the various rocks of the region. There is here a complete ab-
sence of rolled zircon or of other elements that can be considered
as allothigenetic. Specimens from two distinct outcrops of the
green schist associated with quartz veins in the immediate
vicinity of the mine show, along with much rutile that is
evidently of secondary origin, a mixture of the characteristic
elements of the two clays above described, that is to say, per-
fectly sharp crystals of monazite with well rolled zircons, both
in considerable abundance. A polished cross-section of one of
the more massive pieces of this schist shows a banded structure
with alternations of a dark green with an ash-colored material.
The two are so intimately interlaminated that they could not
be completely isolated so as to give perfectly unmixed residues,
but there is an evident concentration of monazite in the green
and of zircon in the lighter-colored portion. Thus both the
character of the residue and the aspect of the specimens indi-
eate that this schist has been produced by the shearing and
metamorphism of a mass of mixed material of which one part,
corresponding to the “horse” of the neighboring mine, is
characterized by allothigenetic zircons and the other, corre-
sponding to the other clay of the mine, by authigenetic monazite.
Thus the veins at Sopa appear to be highly-sheared bodies
essentially identical with the Guia but more completely mingled
with the adjoining rocks.

The observations above recorded indicate the presence in

_the older series of the region of dikes of a rock characterized

by titanium minerals and monazite as primary accessories that
have been sheared and metamorphosed together with the rocks
into which they were injected, and that the schistose and pre-
sumably weak layers thus produced have more frequently than
others been the seat of subsequent injections of pegmatitic
material that passes to pure quartz. In the cases examined

352 O. A. Derby—Association of Argillaceous

the material of the supposed dikes is characterized, with the
single exception of the Dattas rock, by the mineral monazite
which, so far as at present known, may be taken as peculiarly
characteristic of granitic magmas. The original dikes may
therefore be presumed to have been acid porphyries, or basic
phases of such porphyries. If, however, this conelusion is
correct, all eruptives of whatever nature existing in the series
previous to the uplifting and shearing must have suffered a
similar change and afforded planes of weak strata favorable to
the subsequent injection of pegmatites, or of other types of
eruptives. In the overlying series to which the conglomeritic
diamond-bearing deposits examined belong, the quartz veins
observed show none of these characteristics and appear to be
ordinary veins of segregation. The same is also the case with
the very numerous and large veins observed along about fifty
miles of road over a newer series of graywackes, slates and
limestones in the region lying immediately to the west of the
gold and diamond section of the Serra do Espinhaco. On the
other hand, a number of scattered observations indicate that
the veins of the gold-mining districts of the Serra do Espinhaco
present characteristics similar to those of the diamond district
and difficultly reconcilable with the current hypotheses of a
purely segregational origin. In the only one that has been
critically examined, that of Passagem near Ouro Preto, Dr.
Hussak, as already referred, has found characteristic granitic
accessories and contact minerals in a great sulphuret vein, and
as the somewhat similar veins of the Pary and Morro Velho
mines carry garnets in the one case and microscopic zircons in
the other, it may be suspected that they will on close examina-
tion, for which material is unfortunately not at hand, give a
similar result. Dr. Hussak has also shown in his various papers
in the Mineralogical Magazine, on the remarkable association
of minerals in the cinnabar-bearing gravels of Tripuhy, that
these carry monazite and xenotime, the latter of the same type
as in the Diamantina district, along with the curious titano-
antimoniates of which one has been traced to a sericitie schist
very similar in appearance to that above described, and to a
pyritiferous sericitic schist found in place in the immediate
vicinity in which rare prismatic fragments of monazite of the
Sopa type were detected. Lithomarge with other peculiar

clays are, with quartz, characteristic features of the topaz and

euclase- bearing veins near Ouro Preto, and lithomarge is a
constant accompaniment of the peculiar ‘gold and iron-bearing

jacutinga veins (?) in the iron ore beds of the same region.
- Thus the association of argillaceons materials with quartz veins
is wide spread and varied in its occurrence, and the observa-
tions here recorded are suggestive of the necessity of a careful
revision of the current hypothesis regarding the segregational
origin of many of these veins.

Se i al

Rocks with Quartz Veins in Brazil. 353

Both the monazite and the xenotime of the clays and rocks
above described and of the concentrated washings of the mines
to which they have contributed, present peculiarities of form
that distinguish them from the types that have hitherto been
found in place in Brazilian, or other, rocks. The monazite
_ crystals are invariably in elongated prisms with only the pina-
' coid plane developed giving forms that exactly resemble simple
_ prisms of zircon with broken terminations. The extremities
are most frequently broken, but occasionally one or two ter-
minal planes are observed as in the accompanying figure
kindly drawn by Dr. Hussak. The smallness of the extinction

Q&

a (100), b (010), x (101), w (101).

angle (1°-4°) increases the resemblance to zircon, which is so
close that only by chemical tests and the measurement of the
angles (for which material of sufficient size was fortunately
found in the diamond concentrates) could their true character
be determined. The xenotimes are invariably elongated prisms
with pyramidal terminations in strong contrast with the octa-
hedral type, which is the only one hitherto found in the numer-
ous rocks containing the mineral that have been examined.
The crystals when perfect are as transparent and lustrous as
zircons, but are quite sudject to a milky alteration and are
extremely brittle. They rarely sink to the extremely minute
size of the generality of the monazite crystals and show a ten- °
dency to assume macroscopic proportions. Brown opaque
crystals of this prismatic type up to 10 millimeters or more in
length, are quite abundant in the concentrates of the mines
about the village of Dattas, a few miles to the southward of
the localities above described, and there can be no doubt that
they come from quartz and kaolin veins like those of Sao Joao
da Chapada and Sopa. The octahedral type which commonly
occurs in the monazite-bearing granites and porphyries, and
which was actually found in one of the clays from Sopa, is
comparatively rare in the concentrates examined from the Dia-
mantina region, but is abundant and of macroscopic propor-
tions in those from the Lengoes region in Bahia. Perhaps its
tendency to alteration noted in the Sopa clay may explain its
rarity.

354 O. A. Derby—Association of Argillaceous

Since the above was written I am enabled, through the ex-
treme kindness of Profs. Clarke and Hillebrand of the U. S.
Geol. Survey, to add the following very minute and painstak-
ing analysis of the peculiar rock from the Serra do Gigante,
which, as indicated in the above discussion, is the most prom-
ising for giving an idea of the hypothetical original rock
type from which the various schists here described may be
presumed to be derived. At my request special attention was
given to the detection of rare elements of which those of the
cerium group and lithia were, from a preliminary examination,
presumed to be present in an abundance that failed to be veri-
fied. Contrary to expectation, the rock proves to be chloritic
rather than sericitic, though, as 1 am informed, the petro-
graphic examination made in Washington confirmed the opin-
ion formed here that the appearance of the predominant
micaceous element is that of a muscovite mica. Contrary to
expectation also, the rock proves to be basic rather than acid,
and thus the principal difficulty in the way of the attempted
affiliation with the Dattas and Sao Joao schists of diabasic, or
basaltic, aspect is removed. On the assumption, based on the
complete absence of recognizable allothigenetic elements, that
the original rock was eruptive, no affiliation with any known
type of which analyses are at hand can be made.

Analysis of Cyanitic schist, from Serra do Gigante, near Dia-

mantina, Brazil, by W. #& Hillebrand.

Residue insol.

Whole mass. HCl extract. H.SO, ext. in H2SO,4
SiO, 38°32 10°78 14°76 (23-56)
TiO, 4-93 10 20 (4°73)
ZrO 09 trace ? ? (09)?
Al,O, 28°16 10°42 14°77 (13°39)
Fe,0, 2°24 (1°78) 2°24) none
FeO 4:02 (3°21) (4:02 =
(Ni, Co)O 04 (:04) mee
n "16 ” ("16) ener
CaO 32 34 32 2 hehe
SrO trace ? ? mec
MgO 12°04 9°34 (12°04) none
K,O 1 Pe 26 (26) "85
Na,O "16 (03) (-03) 13
Li,O strong trace strongtrace (strongtrace) faint trace
H:O, 105°") "ev5p (55) (55) we
HO, 290°" © 45 :
HO, 310° 9 5°36 6°80 66
H,O, 310°+ 6:79
Pos “47 “47 (47) Bhi
S trace Suit. aly ike
BR trace ? sees tote SOE
100°07 42°64 56°66 43°41

No boron, nor rare earths.

4

Ly

Rocks with Quartz Veins in Brazil. 355

Notes to the Foregoing Analyses.

“The bracketed figures are either assumed or follow from direct
determinations in one of the other columns.

An hour’s treatment with concentrated sulphuric acid is suffi-
cient for extraction of all the soluble constituents. Probably very

rolonged action of hydrochloric acid would have had the same
effect. The silica from the decomposed silicates separated on
rotation in the insoluble form. The lithia belongs to a soluble
mineral, for but a faint trace was found in the residue. The P,O,
is wholly extracted by dilute nitric.acid; it can therefore hardly
exist as monazite. Moreover, careful tests failed to show a trace
of rare earths. Still the lime is quip insufficient to form apatite
with all the P,O..

No formula for the soluble portion can be calculated from the
analyses, but according to Professor Clarke it would seem to
approach in composition an impure earthy chlorite.

The insoluble residue, after treatment by hydrofluoric and sul-
phurie acids, contained besides a large amount of unattacked
rutile, practically only silica and alumina in the proportions re-
quired tor cyanite. Assuming that the alkali of the residue be-
longs to sericite, as indicated by Professor Derby’s letter accom-
panying the specimen, and deducting accordingly for the formula
of typical muscovite (KH,A1Si,O,,), the following percentages
remain after excluding all TiO, as rutile: SiO, 19°62, Al,O, 10°24,
H,O -27. Again assuming that the alumina exists herein as
cyanite and deducting Al,O, and SiO, to correspond, there
remains 13°66 per cent of free silica. This can hardly be opaline
silica, for an hour’s digestion of the insoluble residue with strong
potassium hydroxide solution extracted but a fraction of a per
cent of silica.”

In view of the apparent disaccord in some points of this
analysis with the results above recorded of the physical exam-
ination with batea, heavy liquids and microscope, the latter has
been repeated with special care and on a sufficient quantity (20
grams or more) to secure a reliable result. The presence in
considerable abundance of free quartz and of rutile was fully
confirmed, but no reliable trace of apatite or of zircon could
be found. The former if present at all, which is doubted,
must be in a proportion too small to account for more than an
extremely minute portion of the phosphorie acid and lime, and
the latter is certainly absent. The zirconia found in the analy-
sis must be from some other mineral and presumably from the
soluble constituents of the rock. A mineral closely resembling
zircon, and like it heavier than cyanite, is completely dissolved
out of a mixed residue by sulphuric acid, and in the solution
very satisfactory tests for phosphoric acid and cerium were
obtained, the latter being precipitated as oxalate, and verified
by the Florence method in both a borax and salt of phosphorus

Am. Jour. Sci.--FourtH Series, Vou. VII, No. 41.—May, 1899.
24

856 O. A. Derby—Association of Argillaceous Rocks, ete.

bead. Isolated grains gave also a very satisfactory microchemi-
cal reaction for phosphoric acid and a somewhat doubtful one for
cerium, but no definite reaction for the latter could be obtained
by the blowpipe method, probably on account of the smallness
of the quantity of pure material available or of some disturbing
admixture. On the strength of these observations and of the
close resemblance in form and optical properties with the min-
eral from the Sopa clay, the original determination as monazite
is maintained. The Sopa mineral is undoubtedly a cerium
phosphate and apparently agrees in form with submacroscopie
crystals from the diamond concentrates on which the angles of
monazite were verified. It appears, however, to carry some
titanium and lime, and possibly this peculiar type of monazite
may present peculiarities of composition. The facility with
which the mineral can be isolated from the Sopa schist gave an
opportunity to determine roughly, by washing, its proportion
as about 0:01 per cent. It is apparently much less abundant in
the Serra do Gigante rock and therefore could hardly be ex-
pected to appear in the quantity ordinarily employed in a
chemical analysis. It seems probable also that this rock carries
some other undetermined phosphate.

Commissao Geographica e Geologica,
S. Paulo, Brazil.

—-~ «
a

W. Hl. Hobbs—Goldschmidtite, a New Mineral. 357

Art. XXXIX.—Goldschmidtite, a New Mineral ; by
Wituiam H. Hoss.

THE analyses of Hillebrand and the crystallographical studies
of Penfield which were made in connection with the mono-
eraph of Cross and Penrose* on the Cripple Creek Mining
District have established the fact that the chief telluride ore of
the district is the mineral calaverite, containing about three
per cent, or less, of silver. The presence of this mineral in
the ores had previously been determined by Knightt by
chemical analysis. Mr. R: Pearcet of Denver had, however,
assayed an ore-bearing rock from the Moose Mine, and found
it to contain gold, silver, and tellurium in proportions corre-
sponding with sylvanite, so that he believed this mineral to be
present. From the complex nature of the material assayed
these results can hardly be regarded as conclusive, but to the
writer they would seem to indicate at least the probability of
the presence of a gold-silver telluride, richer in silver than
calaverite, particularly since the output of silver from some of
the mines is too large to be fully accounted for by the presence
of calaverite alone. Dr. Hillebrand says:

“ Notwithstanding that sylvanite has not been identified by
positive chemical or crystallographical tests, the evidence of
Mr. Pearce as to its presence in some portions of the district at
least is entitled to consideration” (I. ¢., p. 133).

Source of the Material examined.

The material which I have studied was kindly given me for
examination by Mr. Gustavus Sessinghaus, a graduate of the
Columbia School of Mines, and now a graduate student of the
University of Wisconsin. It was obtained from the Gold

-Dollar Mine in Arequa Gulch, in the extreme southwestern

portion of the Cripple Creek Mining District. This mine
is located on the northwest quarter of Section 31. (Cf.
Plate VII, J. ¢.) It is near the flank of Grouse Mountain.
The material consists of a single small hand specimen, the
matrix of which is a breccia in which fragments of granite
are conspicuous. Comparison with the gneissose granite of
Grouse Mountain shows that the enclosures in the rock have
probably been derived from that source. The cementing mate-
rial of the breccia resembles the phonolite of the region, with

*Geology and Mining Industries of the Cripple Creek District, Colorado; by
Whitman Cross and R. A. F, Penrose, Jr. 16th Ann, Report, U. 8. Geol. Survey,
1894-5, Pt. II, pp. 1-209.

t Proc. Colo. Sci. Soc., Oct. 1, 1894. t Ibid., April 5, 1894,

358 W. H. Hobbs—Goldschmidtite, a New Mineral.

which it probably corresponds in composition. The fragments
of the breccia are, however, only loosely consolidated, so that
the specimen is crossed by numerous ir repular cracks, the walls
of which are coated with chalcedony.

The erystals of the mineral which is here eaeeed are

firmly attached to the chalcedony, so that great difficulty was.

experienced in removing them without fracture, particularly
as they are possessed of a perfect cleavage. These crystals
have a rather long columnar habit with some tendency also to
elongated tabular forms with reference to a plane parallel to
this axis, and in one or two cases they were observed to have
the arborescent forms which have given the name “ Schrifterz ”
to the mineral sylvanite, and have been explained by twinning.
The largest individual observed was only about 5™™ in lengt
and those which I have succeeded in separating from the chal-
cedony of the walls would average hardly 2™™ in length, with
a thickness perhaps a third or a fourth as much. Examination
with the lens showed these crystals to be considerably modified
and without striation or other distortion, and, except for their
minute size, they are admirably suited to ‘measurement. The
have a perfect cleavage following the plane of their tabular
development. They are seen to be often twins, the plane of
twinning following the columnar axis in a direction normal to
the cleavage. They are quite brittle and have a hardness of
about 2, since they will just scratch the surface of selenite.
The amount of material ‘was not sufficient for a satisfactory
determination of the specific gravity, but the mineral’s compo-
sition as recorded below, and the comparison of it with syl-
vanite and calaverite, make it probable that the specific gravity
is very near to 8:6. The luster of the mineral is bright metal-
lic and the color a silver white. The streak is dull, grayish
black.

Chemical Composition.

Examined on charcoal before the blowpipe, the mineral
readily fuses surrounded by the light bluish green flame which
is characteristic of tellurium, leaving on the coal a white coat-
ing of tellurium oxide. A yellowish white button of gold and
silver results from the fusion, as in the cases of sylvanite and
calaverite, though it is much less yellow than the button
obtained from the last mentioned mineral. No antimony, sul-
phur, or selenium could be detected.

A portion of the specimen was broken up and by careful
and laborious picking out of the erystals and fragments of crys-
tals about one-tenth of a gram of nearly pure mineral was
obtained, which was analyzed and found to contain gold, silver,
and telluriam, in the following proportions, the tellurium
being estimated by difference :

Ps

.
;

yas

W. H. Hobbs—Goldschmidtite, a New Mineral. 359

Percentage composition. Ratio.

[2 /), Looe . 81°41 "1596

Yes er 8:95 08351

ule 5; Ses are en (59°64) 4771
100°00

These proportions correspond almost exactly with the
formula Au,AgTe,, and show that the mineral is a new species,
which I propose to call Goldschmidtite, in honor of my friend,
Professor Victor Goldschmidt, of Heidelberg, the inventor
of the two-circle goniometer and the author of Index der
Krystallformen and Winkeltabellen.

Below are given in parallel columns the estimated and the
theoretical proportions of the constituents of this mineral on
the basis of the formula Au,AgTe,:

Calculated. .
Me te (59°64) 59°95
715 | PaaS A apa AS alt 31°44
Ag Baty SY4 Sted Foye 8°95 8°61
100°00 100-00

Relationships with other Tellurides of Gold and Silver.

The mineral sylvanite contains the same elements as Gold-
schmidtite but in the proportions represented by the formula
AuAgTe,* The mineral calaverite differs in being richer in
gold and is not generally credited with a definite formula,
becanse of slight variations noted in the content of silver.
The number of analyses of this mineral which have now been
made renders it possible, as it seems to me, to arrive at a more
definite formula than (Au,Ag)Te,, which assumes an isomorph-
ous relation between gold and silver in the compound that
does not exist so far as we know. Comparison of the analyses
of calaverite which are given below brings out their striking
uniformity with the exception of the silver percentage, and
this would seem to indicate that a mineral corresponding to the
one having the highest silver percentage (3°52) may form iso-
morphous mixtures with one that is richer in gold or perhaps
entirely free from silver. Since, however, a large proportion
of the analyses contain silver in amounts approximating the
maximum percentage, it is improbable that isomorphous mix-
tures occur with tellurides much richer in silver. The formula
Au,AgTe,, seems best to represent the composition of this
mineral.

* Groth, Tabellarische Uebersicht der Mineralien, 4te Auflage, Braunschweig,
1898, p. 28.

360 W. H. Hobbs—Goldschmidtite,a New Mineral.

Analyses of Calaverite.

F y le Au Ag Total
leC@alifornia, Genth*) 2. Yo .0 sno aee . 55°89 _40°70 3°52 100711
Ds C fee ee Ct eae (56°00) 40°92 3°08 100°00
3. Bowlder Co., pelo: Genth] sca 57°67 40°59 2:24 100°50
A. S-bibdahe 57°32 38°75 3:03 99°10
By Cripple Creek, Colo. Hillebrand||. - 57°60 89°17 3°23 9100200
6: q._ 57:40 40°8355- 77a 0
ie * o cs ee = LOU) eet "90 100°00
Calaverite of formula Au,Ag3Te,,--. 57-40 39°08 3°52 100-00

Comparison of the minerals calaverite, Goldschmidtite, and
sylvanite brings out the fact that Goldschmidtite is, as regards
its composition, intermediate between calaverite ae sylvanite
and almost exactly half way between them. Together the
three minerals represent an homologous series quite analogous ©
to that which has been determined for the minerals of the
humite group by Penfield and Howe.t+ This will be clear
from the following table :

ie Au Ag

= aaa ==) (SS SSS OS SSS

Per cent. Ratio. Per cent. Ratio. Per cent. Ratio.
Calaverite .... 57°40 "4544 39°08 "1987 oO! "03827
Au pacelic. = Bear Op B= 4536) (234 = = 1944) 3= 0363)
Difference See ae ARH "0256 7°54 "0388 5°09 "0473
Goldschmidtite 59°95 °4800 3144) Sooo 8°61 -0800
Ae ert E88 22204752) (22 = 11584) (4,1 = :0799)
Differenceys22. 22°20) 4:0176 6°99 :0336 4°79 :0444
Sylvanite Meee =e Lo "4976 24°45 °1244 13°40 °1244
AuAgTe, eee (23 = 4968) —— = 1235) (434 == 1235)

From the above table it is seen that the common difference
between consecutive members of the series in ascending order
is a gain of Ag,Te and a loss of Au}. For purposes of com-
parison the formule of the three minerals should therefore be
written :

Calaverite .2).. 22208 AuzAo3Te, .
Goldschmidtite ._--- Aut2Agit Te...
Sylvanite...._.-..... AutjtAgi7Te |

The next member of the series should have the formula:
Aut Ag22Te,,

* This Journal, xly, 1868, p. 314. + Ibid.
t Am. Phil. Soc., xiv, 1874, p. 229. S$ lbid.; xvii, 180 7, apane
|| 16th Ann. Report U.S. G.S., Pt. II, p. 133. { Ibid. ** Thid.

Note.-—The mineral krennerite is a telluride of gold and silver richer in silver
than sylvanite, but the wide and remarkable variations in the proportions of its
constituents cause it to be regarded as of very indefinite composition.

t+ On the Chemical Composition of Chondrodite, Humite, and Clinohumite.
This Journal, xlvii, 1894, p. 188-206,

W. H. Hobbs—Goldschmidtite, a New Mineral. 361

which is a close approximation to Au,Ag,Te, The mineral
krennerite contains silver in about this proportion (19-44 per
cent), but the proportions of gold and tellurium do not corre-
spond.

Crystallography of Goldschmidtite.

As already stated, the chief difficulty encountered in measur-
ing the crystals has been their minute size and the conse-
quently small amount of light reflected from their faces. A
number of the better faces afforded single images of the signal ,
so that in the prism zone, where the faces were largest, con-
secutive readings from the same angle showed variations of less
than a minute of arc, and on the best crystal the four similar
angles between prism and pinacoid were respectively :

61° 42

61 46

61 464

61 474

Considerable confidence is therefore placed in the correct-
ness of the constant which was determined from the angles in
this zone, while the values of the other constants are correct
only within limits of error which are sufficiently indicated by
a comparison of observed and calculated values of readings of
interfacial angles. Five crystals have been completely meas-
ured and particular angles have been measured also upon other
erystals.

The symmetry of the mineral is easily determined as mono-
clinic, and it is probably also clinohedral. The principal zones

of the crystal are those of the axes ¢ and 4, a single clinodome
and a pyramid which was too small to be determined being the
only faces not included in these zones. Owing to the frequent
absence and the small development when present of the basal
pinacoid, the angle 8 was determined from the mutual inclina-
tions of the positive and negative unit orthodomes and found
to be 89°11’. The axial ratio a: b:¢ is 1°8562:1:1:2981. The
crystal forms observed number twenty-two, and are as follows:

Zone of &.— a (100) 77; b (010) 7-2; m (110) ZT; f (210) 72;
g (310) @3 5 ¢ (370) 74; 7 (130)73.

Zone of b.— y (508) — 2-7; s (101) — 1-7; m (201) — 2-7; 7 (708
— 47; w (401) — 47; g (801) — 87; @ (10°01) — 10-7 5 v (35-071
— 35-1; 8 (101) 17; N (201) 2-7; W(401) 4-7; X (16-0°1) 107;
Z (14'0'1) 14-7; ¢ (001) O.

Zone of d.—k (032) 3-2:

The more important angle readings from which these deter-
minations were made are as follows:

362 W. H. Hobbs—Goldschmidtite, a New Mineral. —

Obs. Cale. Diff.
Ce ei eT O) ne Se, (1 10) Gla
bsg me aONO), s \(110)) ) See eens ae =a
Oe f (200), + (210) .. 40a psoas —9
@ 2g 100 (310) 81.55." 31 4S aaeeen
Dew. Ufone ts: “(180) 9. 57.) 10 LIers
Devi wi aON0) 032 26.49... 9% deans
a: y (100) 3 (508) 65 7%) 65 oe
Geis (100) (101) 54 57 54 29 +28
arn (LOO). ys A(20iy) 34 13 35 17 —64
a@:r ° (100) : (708) 31 40 3a ees
Dae 0 (100) 3401) DS 1k 19 35 —25
UIE) 100) (801) LORS LO —3
a: (100) : (10-01) 7 44 Bh) eo
a 2) (100): (85-00) 2 99 2 20 49
_@:e¢.. (100) ; (001).°. 89 25 . “69 Stimeeee
b:c¢ (010) (001) 89.59 90 0 4s
mite es (O10) 370). 19.534. dee =
ae 1S. 4100) (101) 55 35%
ag = No (100) (201) 34 58. 36070 teumee
a: W_ (100) 401) 19 18° (19 46) 2 eee
aa ee LOO) 2 (100) 7 48 8° 9) aoe
ae Fe M00) = 14-019 5 49 5 44 45
been’ (010) (032) . 96 49. Of -alen egies

] Ze 3

IN
ION
1 \
Lhe Rae
| b=
! 1
t
| }
|
| 4
| \
| t
'
i}
!

Of the five crystals measured, crystal I (fig. 1) is bounded
by the forms a, b, m, f, 9, l, ¥, 9, % ¥ and X. Crystal
-2) is bounded by a, 6, m, f, w, n, 8, Z, W, and N. Crystal II
(fig. 3) exhibits the combination a, 6, m, c, w, n, X, N, and 8.

W. H. Hobbs—Goldschmidtite, a New Mineral. 363

Crystal IV (fig. 4) is doubly terminated and represents a simple
twin with the twinning plane the orthopinacoid. One of its
individuals (the front one in the figure) displays the forms a,
b, m, f, t, n,.8s, N,S,and %. The other individual differs in
having only one positive orthodome (N) and the form w in

4 place of n. Crystal V (fig. 5) is also a twin erystal, on the

front individual of which are developed the forms a, b, m, c¢, k,
and s. The other individual exhibits in addition 7 and lacks
the basal pinacoid. A small pyramid lying between # and s I
was unable to determine. All the crystals show a remarkably
perfect cleavage following the plane of symmetry.

Crystallographical Affinities of Goldschmidtite and Sylvanite.

‘As might be expected from its chemical relationship to sy]l-
vanite, the mineral Goldschmidtite exhibits affinities also in its
erystal development. Both minerals have monoclinic sym-
metry and somewhat similar
forms are developed upon
them. Ten of the twenty-
two forms discovered on
Goldschmidtite have repre-
- sentatives on sylvanite. Be-
_ low are given in_ parallel
columns the crystallograph-
ical constants of the two
minerals and some of the
angle readings between faces
of corresponding forms. It
will be noted that the dif-
ferences are greatest in the
prism zone, where the largest
faces of Goldschmidtite make the error of reading compara-
tively small. The orientation of sylvanite is that of Schrauf,
which is now generally adopted.

The axes @ and ¢ are each about one-seventh longer in Gold-
schmidtite than in sylvanite. It is to be regretted that satis-
factory material is not available for the crystallographical study
of calaverite. That studied by Penfield* was so poor that he
was unable to determine the crystallographical constants. He
thinks the mineral is probably triclinic, with its angles some-
what similar to those of sylvanite. It lacks the pinacoidal
cleavage of that mineral but is observed in twins according to
the face of (101). When suitable material is available for:
study it will be interesting to see whether relations between
the lengths of the crystallographical axes and the proportions

4. 5.

#1. csp. kod:

364 W. H. Hobbs—Goldschmidtite, a New Mineral.

of its constituent elements, can be discovered, as has been done
in the case of the minerals of the humite group.

Goldschmidtite 3 Sylvanite
Au, AgTe, AuAgTe.,.
Monoclinic Monoclinic
B= 89° 11 B= 2973s
Geto 20 BO ae, O53 "Og ae
178562 "21 2 igen 16339. 3) Seas
Perfect cleavage (010) Perfect cleavage (010)
Common twinning plane (100) Common twinning plane (101)
Columnar axis, ¢ Columnar axis parallel to (101)
Angles Angles Diff.
am 61° 41’ 58° 32). cece
bs sm 28 18 31 28 —3 10
fe ee 42 42 39 15 +3 27
Stee 54. 57 54 57
Dose 55 35 54 52 + 43
Tach, 34 13 35 48 =i se
1508 20 43 TOk ei +1 34
as IN 34 58 30 16 — 18
CeO 89 25 89 35 orga
SaccauN: 20 37 19 36 alas aoa
Conclusions.

1. Goldschmidtite is a new mineral species of composition
represented by the formula Au,AgTe,.

2. It occupies an intermediate position exactly half way
between calaverite and sylvanite.

3. Calaverite, Goldschmidtite, and sylvanite form together
an homologous series, analogous to that of the: minerals in the
humite group, the common difference being the addition of
Ag,Te and the subtraction of Au in passing from any one of
the series to the next above it.

4, Goldschmidtite has monoclinic symmetry with 8 89° 11’,
and @ : 6:6, 1:°8562:1:1:2981. It exhibits no less than
twenty-two crystal forms, most of which are in the zones of
c and 0. |

5. Crystallographically as well as chemically Goldschmidtite
shows affinities with sylvanite, ten forms being common to the
two minerals, though they are referred to axial unities of
which @ and ¢ are each about one-seventh longer in Gold-
schmidtite than in sylvanite.

University of Wisconsin.

eR a he a ee ee ee

~

*
6
of

+

Clarke and Darton—Hydromica from New Jersey. 365

| ; Art. XL.—On «a Hydromica from New Jersey; by F. W.
CLARKE and N. H. Darton.

WB5ILE studying the Jura-trias formation in New Jersey,
one of us (Darton) found in an old “trap” quarry at Rocky
_ Hill a hydromica, which occurred under such novel conditions
that it appeared to be worthy of investigation. It is found in
veins of calcite, mainly as a thin coating, and adjacent to the
diabase of the vein walls. The latter consist of more or less
decomposed rock, of which the principal product is a soft,
dark-green chloritic material. In portions of the vein the
mica extends down the cleavage planes into the masses of cal-
cite. A considerable amount of the calcite was thrown out
during the quarrying operations, but only a portion of it is
covered with the mica. This portion presents the appearance
of having been coated with bronze paint.

The mica occurs in minute flakes thinly matted together.
ts color is golden bronze, although some portions are slightly
greenish. The mineral is soft, and thinly foliated. Under
the microscope it exhibits no definite crystalline form ; and its
_ optical properties, although not distinctive, suggest biotite. It
appears to be biaxial, but with a very small axial angle, and it
is pleochroic. When heated, it does not exfoliate. It fuses
before the blowpipe, at a moderately high temperature, to a
dark colored bead. The specific gravity was not determined.
It is readily decomposable by hydrochloric acid. The analysis,
by Mr. George Steiger, of material not free from calcite, is
subjoined. In the second column of figures the reduced
analysis is given, titanic oxide and calcite being thrown out,
soda recalculated to terms of potash, and the whole adjusted to
100 per cent.

Found. Reduced. Ratios.

“ee 32°72 40°24 ‘671
Cea). i: mee eae
CL GE Sarai yy! 10°34 101
en 9 Shey tae Se aaa 19:99 24°57 154
Or es ee NT eS 4°24 The S | 072
Mere Nee ewig Slee 10°30 aoe itt.
ae Vue ot. 5 5°51 6°78 166
ee a... BE 2°20 024
Lk I EES eae 63 Poe we

Ce ae 8-21 ee za
PGF at 100" 3a. 2. 247 3°03 168
H,O above 100° ...--- 6°22 7°63 424

366 Clarke and Darton—Hydromica from New Jersey.

This gives, as an orthosilicate, the formula
(KH),,(MgFe),,(AlFe), (SiO,),,. 28H,O.

It is evident, from these data, that the mica is one which hee

been largely, but not wholly, ‘altered to a vermiculite; the. :

latter term indicating a mica in which potassium has been
replaced by hydrogen, and which has taken up water of crys-
tallization. So far as the analysis goes, the condition of the
water is uncertain; for it was determined in two fractions
only, at and above 100°, whereas more fractions are needed for
accurate diagnosis. Some crystalline water may be retained
far above 100°, so that the loss above that temperature includes
part of this fraction plus all the water of constitution. Apart
from this uncertainty the ratios reduce easily in terms of the
mica theory to the following molecular mixture :

_ASi0, = KH, _Si0O,=H, _Si0, = R’H
3Al—Si0, = =, eR SRO: —R 4 sR SiO, =R’H 8 aq.
SiO, = Al SIO, RM “SiO, =e

ealen ins with the atomic ratios Al: Fe’” :: 2:3, and Fe”:
:3:7, we have the Tolleyume comparison between ual ye?
ae theory :

Reduced. Calculated.
SiO te iis 2 40°24 40°55
PRO Meas 4 hl Cae 10°34 10°51
Wee ou, 2. cee 24°71
MeO ii Sob 52. eed ee 5°21
NIG OF eae 6°78 6°75
AO Gay che es 2°27
EL OPNOOT. a. bess ee 3°03 Crystalline 4°34
EU Ovabove 100°) 25.2) 7°63 Constitutional 5°66

100°00 100°00

In short, the mica consists of muscovitic and phlogopitic mole-
cules in the ratio of 9:5.

9R’”, (Si0,), i
SR” (SiO, y RR” , 3 aq.

The mineral is evidently an eal mica, differing widely
from any other hitherto described. Its very high proportion
of ferric oxide is its chief characteristic, and suggests a ferric
muscovite as one of the antecedent, unaltered molecules. Such
a muscovite is theoretically conceivable, but is not actually
known.

Laboratory U. S. Geological Survey, Feb. 18, 1899.

C. Palache—Powellite Crystals from Michigan. 367.

Art. XLI—Powellite Crystals from 2 Ade by
CHARLES PALACHE.

PoWELLITE was first described by Melville* from a speci-
men from the “Seven Devils” Mountains, Idaho, where it
occurred associated with copper ores and garnet. This material
was well crystallized and Melville was able to establish its
tetragonal character and the forms ¢(001), p (111), and ¢(101).
The isomorphism with the scheelite group suggested by the
close agreement between the axial ratio and angles of powellite
and scheelite was not fully shown, thongh the author says:
“Small rudimentary planes appear on some crystals at the

lower portion of the combination edges (111) (101), thus sug-

gesting hemihedrism as in scheelite. Indeed the curved sur-
face which often replaces these edges, giving the appearance of
fused edges, adds greatly to the evidence in favor of this sup-
position.”

A second occurrence of powellite, in the South Hecla Cop-
per Mine, Houghton County, Michigan, was described by
Koenig and Hubbard.t Their material was poorly crystallized
and added nothing to our crystallographic knowledge of the
mineral. It differed from the Idaho mineral in its paragenesis,
occurring with native copper and epidote, and in its composi- |
tion, having only about 14 per cent of its molybdic acid
replaced by tungstic acid, as against about 10 per cent in the
first occurrence.

Powellite continues to be found at the Michigan locality,
though in small quantities and for the most part in massive
form. Through the kindness of Mr. John T. Reader, of
Calumet, through Dr. L. L. Hubbard, two particularly fine
erystals came into the writer’s hands and their description
follows.

Both crystals are removed from the matrix upon which they
grew. ‘They are perfectly free from impurity of any sort and
the specific gravity of the larger one could be determined with
accuracy by weighing on the dee balance. Two deter-

_minations gave 4°358 and 4°3

The crystals are pale bluish 1 or een in color, translucent, with
subadamantine luster on the er rystal faces and a oreasy luster on
fracture surfaces.

The larger of the two crystals is a centimeter in height and
has one end complete, but the faces are somewhat uneven; the
smaller one is half a centimeter in height and diameter and has
both ends developed though neither is quite complete. The

* This Journal, vol. xli, 1891, p. 138.
+ This Journal, vol. xlvi, 1893, p. 356.

368 C. Palache—Powellite Crystals from Michigan.

quality of its faces is sufficiently good to insure accurate
measurements. . 7

The measurements were made on a two-cirele goniometer
(Goldschmidt model) and the observed angles therefore appear
in the form of the angular codrdinates, ¢ and p, as used in the
Winkeltabelle of Goldschmidt.

The forms observed were:

p(lll) ‘e(101).°A(133) © f(8-1111) = 2iGSs ie ee
and the following table shows the observed and calculated
anges. |

Symbol. ~~ Observed angle. Calculated angle.

= naar tit Gan =a Se > a ee mies

Letter. Miller. Gold’t. @ p o p observed.
€ 011 01° 00°00’ | 57° 042" 00° 007 Sa um 2
p Teh bat 1 44 59 65 24 45 00° 65 24 2
h 153 +1 18°28: 58 254° 18 26 ee 2
goto ded 8 15 08 “58°00” 15 1a aeaae 3
k 155 +1 1 AO oir 11 18335433 5m ike
@ dill 41 5 Le aa Oe 5 Lb oaal 1

As shown by the accompanying figure, the pyramidal hemi-
hedrism is strongly emphasized by the faces of 4(183) and
j(811:11), the first especially being broad and of perfect
quality. The other two hemihedral forms, #(155) and
(1:11:11) were very narrow faces in the zone, and as they were
each observed but once they must be
considered as doubtful and were there-
fore not introduced into the figure.
There is a tendency for these planes to
form a striated or curved surface between
the forms py and e such as Melville
observed, but there were two zones on
the crystal in which the faces were
sharply enough bounded to give good
measurements.

The measurements lead to the same
axial ratio as those made by Melville and
the isomorphism of the species with the
scheelite group is thoroughly established.

With the crystals of powellite sent by Dr. Hubbard was
another specimen of the mineral of quite a different appear-
ance. It consists of a rounded mass of powellite about two
centimeters in diameter, upon which are a few small fragments
of rocky matrix, minute crystals of bright green epidote and
two or three shreds of native copper. On the broken surfaces
the powellite shows an imperfect cleavage, seemingly parallel
to e101). The central part of the specimen is of the same

C. Palache—Powellite Crystals from Michigan. 369

_bluish-green color as the crystals, but the surface layer to the
depth of about three millimeters appears quite black and
opaque in reflected light and when examined in thin splinters
is seen to be colored a deep prussian blue. The boundary
between the two colors is not sharp but the darker portion
shades off rapidly into the lighter part.
The dark-colored material is crystallized in a number of small
‘individuals, all in parallel position and showing approximately
_the same forms as the crystals before described but with
_ greater proportional development of the forms ¢ and A.

An attempt was made to separate some of the darker mate-
rial for chemical examination, it being assumed that the lighter
colored part did not differ from the similar material analyzed
_ by Koenig and Hubbard; but it proved impossible to obtain a
sufficient amount for any satisfactory tests. A small amount
_ (051 gram) was however picked out in which there was a very
slight admixture of light-colored substance and a few minute
grains of epidote, and its specific gravity was determined to be
4214. This determination is not wholly reliable on account
_of the impurities and the small amount of material, and it is
probably too low; but it seems to the writer fairly certain that
the specific gravity of the dark material is less than that of the
light (4355). The specific gravity of the pure calcium molyb-
_ date has been calculated to be 4:267 (Melville), a value which
is rapidly raised by even small quantities of the tungstate. It
seems probable, in view of these facts, that the dark blue
_ material represents an occurrence of almost pure calcium
molybdate. It is much to be hoped that more material will
be found which will permit the chemical proof of the exist-
ence of this end member of the scheelite-powellite series.

Mineralogical Laboratory,
Harvard University, February, 1899.

370 Gooch and Havens— Volatilization of Iron Chlorides

‘

Art, XLII.—The Volatilization of the Iron Chlorides in
Analysis, and the Separation of the Oxides of Iron and

Aluminum; by F. A. Goocw and FRANKE STUART 4

HAVENS. : .
[Contributions from the Kent Chemical Laboratory of Yale University—LXXXI.]

Ir is well known that metallic iron is easily acted upon by
an excess of chlorine at moderately elevated temperatures with
the formation of ferric chloride, and that the product of
the action of hydrochloric acid gas upon the metal is ferrous
chloride. Out of contact with air, or moisture, both chlorides
may be volatilized at appropriate temperatures—the ferric
chloride below 200° O.; the ferrous chloride at a bright red
heat. If water vapor, or oxygen, or air be present during the
heating, both chlorides are partially decomposed with the for-
mation of non-volatile residues, ferric oxide or ferric oxy-
chloride.

Analytical processes involving the volatilization of iron at
temperatures more or less elevated, in an atmosphere of chlo-
rine or hydrochloric acid, have been the object of considerable
attention. Thus, Fresenius,* Drown and Shimer,t and Watts,t
have heated crude iron in chlorine to remove the metal and

leave the non-volatile constituents; and Sainte-Claire DevilleS

has employed hydrochloric acid to volatilize iron from mix-
tures of that metal with alumina (obtained by heating the
mixed oxides of iron and aluminum in hydrogen according to
Rivot),| exposing the mixture, contained in a porcelain boat
and placed within a porcelain tube, to the bright red heat of a
charcoal furnace—an operation which was bettered by Cooke’s4
use of a tube of platinum instead of the porcelain tube and a
gas blowpipe in place of the charcoal furnace. Sainte-Claire
Deville** showed, further, that ferric oxide may be converted
to ferric chloride and volatilized at the heat of the charcoal
furnace if the current of hydrochloric acid is sufficiently rapid ;
but the curious effect was observed that in a sufficiently limited
current of the acid no chloride whatever was volatilized, while
the amorphous oxide was converted to the highly crystalline
oxide of the same composition—a phenomenon which gave
rise to a theory of the natural formation of specular iron in
volcanic regions.

Quite recently, Moyert+ has made record of an unsuccessful
attempt (in the course of experimentation upon the volatility

* Zeitschr. fiir Anal. Chem., iv, 72. + Jour. Inst. Min. Eng., viii, 513.
t Chem. News, xlv, 279. § Ann, de Chim. et de Phys. [3], xxxvili, 23.
|| Ann. de Chim. et de Phys. [3], xxx, 188. [This Journal [2], xlii, 78.
** Compt. Rend., lii, 1264. tt Jour. Amer. Chem. Soce., xvii, 1029.

in Analysis, and Separation of Oxides of Iron, ete. 371

of certain chlorides at comparatively low temperatures) to con-
vert ferric oxide completely to ferric chloride by the action of
gaseous hydrochloric acid at about 200° C. At this tempera-
ture the greater part of the iron sublimed, but a residue
remained, which, volatilizing neither on long heating at 200°
nor upon considerable elevation of the temperature, proved
upon examination to be ferrous chloride. In the experiments
to be described we have acted with gaseous hydrochloric acid
upon ferric oxide made by igniting the nitrate prepared from
pure iron deposited electrolytically by high currents passing
between electrodes of platinum in astrong solution of ammonio-
ferrous sulphate. The oxide, contained in a porcelain boat,
was heated within a roomy glass tube over a small combustion
furnace. The hydrochloric acid (generated by dropping sul-
phurie acid into a mixture of strong hydrochloric acid and
salt, and dried by calcium chloride) entered one end of the
tube and passed out at the other through a water trap. In
early experiments a high-reading thermometer was inserted
through the stopper in the exit end of the tube so that its bulb
was above and immediately adjacent to the boat carrying the
oxide. In this way the actual temperatures of the vapors
about the boat were fixed with considerable accuracy ; later,
after a little experience in gauging the effect of the burners, it
was found that the temperatures could be regulated very
closely without actually depending upon the thermometer.
We found, as did Moyer, that ferric oxide, submitted to the
action of dry hydrochloric acid gas, volatilizes partially as ferric
chloride at low temperatures—180° to 200° O.—leaving ulti-
mately a crystalline'residue which does not change visibly when
heated for an hour or two at 200°, or even at 500°, in the pure
dry acid. According to our experience, this residue is generally
slightly reddish or salmon-colored; but sometimes, especially
after a second heating, the boat having been withdrawn from
the tube or exposed to the atmosphere (and so to moisture), the
residue is white. When it is white it dissolves in water,
- yields the characteristic reaction for a ferrous salt with potas-
sium ferricyanide, gives no reaction with potassium sulpho-
cyanide, and upon treatment in weighed amount with potassium
permanganate destroys the amount of that reagent theoretically
required for its oxidation upon the supposition that it is fer-
rous chloride. The slightly colored residue when treated with
water yields a solution showing the reaction of a,.ferrous salt
only, but when treated with hydrochloric acid and then tested
shows the presence of a trace of iron in the ferric condition.
Doubtless the coloration of the residue is due to an included
trace of ferric oxide or oxychloride, which after exposure of
the containing crystals to slight atmospheric action, is more

Am. Jour. Sci—FourtH Series, Vou. VII, No. 41.—May, 1899.
25

372 Gooch and Havens—Volatilization of Iron Chlorides

easily reached in the second heating by the gaseous acid. The
amount of residue is somewhat variable, but approximates
under the conditions of our work to from five to ten per cent
of the oxide taken: thus, in one typical experiment 0-1 grm.
of ferric oxide left a residue which (withdrawn. after cooling)
weighed 0°0115 grm.

The greater portion of the ferric oxide volatilizes when sub- -
mitted to the action of the gaseous acid at 200° quickly and
abundantly in the form of the greenish vapor of ferric chloride,
and if the operation is interrupted at this stage the residue
which remains is nearly black, insoluble in water, slightly
soluble in cold hydrochloric acid, and readily soluble in hot
hydrochloric acid with the formation of ferric chloride. It is
probably something analogous to the oxychloride which
Rousseau* identifies as the product of the action of water
upon ferric chloride at 275° to 300°. This dark residue yields to
the action of the hydrochloric acid at 180° to 200° only slowly,
but ultimately only the residue which is essentially ferrous
chloride remains; thereafter little volatilization oceurs within
the range of temperature of our experimentation—200° to
500°. |

It is obvious that a reduction of iron in the ferric condition
to iron in the ferrous condition takes place under the conditions
of our work, and it is difficult to see how this can occur other-
wise than by the direct dissociation of ferric chloride under
the low partial pressure conditioned by the brisk current of
hydrochloric acid gas. The temperature of formation, 180° to”
200°, is far below that at which such dissociation is supposed to
begin. Thus, Gruenewald and Meyert found, after cooling,
no evidence of the dissociation of ferric chloride which had
been heated in the Victor Meyer vapor-density apparatus to
448° in contact or partial mixture with nitrogen ; but ten per
cent of the residue obtained by heating to 518° was in the
ferrous condition. Friedel and Crafts,t however, did see erys-
tals of ferrous chloride at 440° on the walls of a Dumas con-
tainer filled with the vapor of ferric chloride and nitrogen, the
former exerting a partial pressure of 0°75 ; while ferric chloride
volatilized into an atmosphere of chlorine without evidence of
dissociation. It seems rather surprising, therefore, to find so
large a percentage of dissociation as that shown in our experi-
ments at a temperature so low—180° to 200°. Curiously, too,
we find, on.repeating the experiment of heating ferric oxide in
gaseous hydrochloric acid, that if the temperature of the oxide
is 450° to 500° when the brisk current of acid begins to act, the
whole mass of oxide is converted and volatilizes without resi-

* Compt. Rend., exvi, 118. + Ber. d. d. Chem. Gesell., xxi, 687.
¢{ Compt. Rend., evii, 301.

in Analysis, and Separation of Ouides of Iron, etc. 378

due. It is hardly to be supposed that the degree of dissocia-
tion at 450° to 500° can be less than that at 180° to 200°, and
atest of the sublimate, after cooling, shows that it contains a
ferrous salt. Plainly, ferrous chloride (formed by dissociation)
has volatilized, and inasmuch as the ferrous chloride constitut-

: ing the residue formed at 180° to 200° does not volatilize in

the hydrochloric acid even at 500°, itis plain that the volatility
of the former is not determined by the presence of the latter.
Apparently, the cause of the completeness of volatilization
must be sought in its rapidity; and this is not an unreasonable
hypothesis, if one considers that an action sufficiently rapid to
keep above the boat an atmosphere of ferric chloride and its
products of partial dissociation, might naturally provide the
very condition which would be effective in counteracting the
tendency of the residue to dissociate before it volatilizes. If
this hypothesis is correct, it is plain that the introduction of
chlorine gas, the active product of dissociation, into the atmo-
sphere of hydrochloric acid ought to bring about the volatiliza-
tion of the residue of ferrous chloride, formed at 180° to 200°,
which refuses to volatilize in the acid alone. As a matter of
fact, we find by experiment that if a little manganese dioxide
is added to the contents of the generator, so that the hydro-
chloric acid may carry with it a little chlorine, every trace of
ferric oxide is volatilized from the boat at 180° to 200°; and
the residue of ferrous chloride found at 180° to 200° when the
hydrochloric acid is used alone is likewise volatilized at the
same temperature, when the admixture of chlorine is made.
These facts, that ferric oxide is completely volatile in hydro-
chloric acid gas applied at once at atemperature of 450° to 500°
C., and at 180° to 200° if the acid carries a little chlorine,
open the way to many analytical separations of iron from sub-
stances not volatile under these conditions. In the experiments
of the following table we have applied these methods to the
separation of intermixed iron and aluminum oxides. The
ferric oxide employed was made, as before, by ignition of the
nitrate prepared from iron deposited electrolytically by a
strong current passing between platinum electrodes in a solu-’
tion of ammonio-ferrous sulphate.* The aluminum oxide was
made by igniting to a constant weight the carefully washed
hydroxide precipitated by ammonia from a pure hydrous
chloride thrown down from the solution of a commercially
pure chloride by hydrochloric acid.t| The hydrochloric acid

* The use of an anode of commercially pure iron wire naturally facilitates the
operation, but in our experience the deposit thus obtained is likely to carry traces
of impurity. In an attempt, too, to prepare pure ferric oxide from the oxalate
thrown down out of ferrous sulphate with all precautions, the material obtained
still held traces of silica, and possibly alumina, amounting to 0°0004 grm. in 0°1
grm. of the oxide.

+ This Journal, II, 346.

374 Gooch and Havens— Volatilization of Iron, ete.

gas was made by dropping sulphuric acid into strong hydro-
- chloric acid mixed with salt, and a little manganese dioxide
was added when the mixture with chlorine was desired. The
experimental details are given in the table.

Fe.0; Al,O3 Al,O3
taken. taken. found, Error. Time. Temperature.
erm. erm. erm. grm. hours. Ce Atmosphere.
0°1000 thi Sie De ake 0°0000 450-500 HCl.
0°2000 Me eae Ben 0°0000 h +

Ce in4

0°1020 0°1015 0°1015 0:0000
0°2145 0°1006 0°1008 +0°0002

OADO GO. Sno De Lhe, OOO
071000 0°1082 0°1082 0°0000
O°'1072 0:1018 0'1015 +0°0002
0°2045 . 0°1032 0°1033 -+0°0001
071050 = 0°1023 071019 —0:0004
0°2008 0:1007 0°1006 —0:'0001

Sas 0°1087 0°1087 0'0000

6¢ “cc

180-200° HC1+Cl.

6<
« 6c
G6 T7

450-500

(74 (4

bt loo bole re mel ico ito leo noes RY BS
Bie

66 (45

The residual alumina tested in several experiments by fusion
with sodium carbonate, solution in hydrochloric acid, and addi-
tion of potassium sulphocyanide gave no indication of the
presence of iron.

The separation of the iron is obviously complete at 450° to
500° when the mixed oxides are submitted at once to the action
of hydrochloric acid gas, or at 180° to 200° when chlorine is
mixed with the hydrochloric acid. Plainly, the extremely
high temperatures employed by Deville are unnecessary if the
mixed oxides are submitted at once to the action of hydrochloric
acid at 450° to 500° without previous gentle heating in the acid
atmosphere. We prefer, however, to use the mixture of
chlorine and hydrochloric acid, not only because the tempera-
ture of the reaction is lower, but because it needs no regula-
tion, while the danger of error arising from the liability of
ferric chloride to dissociate, or from deficiency of oxidation in
the oxide treated, or from mechanical loss due to too rapid
volatilization, is avoided.

A. E. Verrill—New Actinians. | 375

| Art. XLIL— Descriptions of imperfectly known and new
| ~—s-—s Acctinians, with critical notes on other species, V; by A. E.
Verritt. Brief Contributions to Zoology from the Musewm
of Yale College, No. LXIL.

Family Bunopactips. (Continued from p. 218.)

Bunodactis inornata Verrill. Figure 39. :

Actinia inornata Stimpson, 1855.

Bunodes inornata Verrill, Proc. Essex Inst. Salem, Mass., vol. vi, p. 61, [27],
1869, pl. i, fig. 5.

THE type of this species is still preserved, though it has at
some former period been partly dried, and is, therefore, much
contracted, in a hemispherical shape. The tentacles are wholly
retracted. The upper part of the wall is closely covered with
rather large concave suckers, but they rather suddenly diminish
in size and become confused with the deep, vermiculate,
wrinkles on the lower half of the column, though some persist
to near the base.

In section, there are 24 pairs of larger complete mesenteries,
including two pairs of directives. Alternating regularly with
these there are 24 pairs of smaller mesenteries, most of which
are incomplete, but are well developed, with a-small but promi-
nent muscular pennon near the distal end, occupying a quarter
or less of the breadth. The larger mesenteries have strong
median muscular pennons, occupying about a third of their
breadth. All, or nearly all, the septa bear large gonads. The
tentacles, in transverse sections, show a stellate muscular layer,
due to the strong radial supports.

The sphincter muscle is nee) clearly defined, nearly circu-
lar in section.

Hong Kong, at low water mark, in gravel.—Dr. Wm.
Stimpson.

Some Actinians that incubate their eggs externally.

The next four species, described below, are remarkable for
the habit of carrying their relatively large eggs attached to the
body in special pits, where they are retained until hatched and
provided with tentacles. In these cases the egg develops
directly into an actinula. This habit was described by me in
1869 as occurring in two of these species, but it has not
attracted the attention of others to any great extent. Iam
now able to add two other species, of large size, having the
same habit in a still more striking degree. They belong to
the Bunodactide, but differ considerably i in generic characters.

376 A. F. Verrill— New Actinians.

Pseudophellia, gen. nov. Type, P. arctica Ver.

Column much elongated, covered with an adherent duties
much as in Phellia, except near the summit. When the
cuticle is removed the surface is finely papillose and wrinkled,
in alcohol. Below the middle there are several transverse rows

of large pits for carrying eggs. These are deeply excavated in

34. 35.

the wall by a thinning of the mesoglea. Capitulum smooth
Wall thick and firm. Sphincter muscle well developed, cir-
cumscribed. Mesenteries form about 24 perfect pairs with a
few impertect ones. Two siphonoglyphs and two pairs of
directives in the type. Tentacles tapered, in two or more
rows. |

A. f. Verrill—New Actinians. OLE

Pseudophellia arctica Ver. Figure 34 (type).

Phellia arctica Verrill, Proc. Essex Inst., vol. v, p. 328 [14], 1868; Trans.
Conn. Acad., i, p 490, 1869 (not of Danielssen*).

The type of this species is still preserved. The original
description is pretty complete as to the external appearance,
but a typographical error occurred, in the measurement of the
egos. They are ‘05 of an inch in ‘diameter (not 5). When I
originally described them I did not feel certain that they were
not parasitic structures, but subsequent discoveries have
removed this doubt. The mouth has two strong siphonoglyphs
and about 18 folds on each side. The stomodzum is strongly
plicated. There are about 24 nearly equal pairs of perfect
mesenteries, with a few small ones of the fourth cycle. The
longitudinal muscles are rather thick and strong on the per-
fect mesenteries.

The tentacles are rather large and moderately long; they are
only partly retracted in the type, but the sphincter muscle is
well developed, so that they are probably capable of complete
retraction and concealment. In section the sphincter muscle
has an ovate outline. The wall is rather thick and firm with
a thick mesoglcea, except in the’ capitulum, which is smooth
externally and forms a distinct submarginal fold and small
fosse. Circular muscular layer of the wall is well developed.
The cuticular layer is thick and soft, and often nearly or quite
conceals the eggs when enclosed in ‘the pits, which are largely
excavated in the mesoglea.

Fpiactis prolifera Ver. Figure 25.

Trans. Conn. Acad. i, p. 492, 1869.

Epiactis fertilis Andres, op cit., p. 363, 1884.+

The figures now given are from one of the original types.

The largest specimens seen have about 96 tapered tentacles ;
they can be entirely retracted and concealed.

The wall, in section, is moderately thick and firm, with the
different layers very distinct ; the ectoderm is rather thick and
close; muscular layer well defined and continuous. The
sphincter muscle is large, clearly circumscribed, ovate in sec-
tion, essentially endodermal, but not so much detached from the
wall as usual in Lunodactide. There is a distinct collar and
fosse, in contraction. ‘The mesenteries are regularly hexamer-
ous, in four cycles, with some small ones of the fifth; 12 or

* The Phellia arctica Dan. (Actinida, p. 54, 1870) is a true Phellia and may be
called P. Danielsseni, to avoid confusion with our species.

+ Andres changed the specific name on account of its prior use for Gonactinia
prolifera Sars, but that belongs not only toa different genus, but to a widely dif
ferent family, so that the change of name was not necessary.

378 A. E. Verrill-—_New, Actinians.

more pairs are perfect and fertile. Their longitudinal muscles
are not very thick, but cover most of their breadth.

The stomodeum is strongly plicated and has two large
siphonoglyphs. The ovaries, in some cases, contain large eggs,
like those carried on the outside. The egg-pits are not so deep
as in the preceding three species and at first they are confined
to the ectoderm, but as the embryos develop the pits become
larger and deeper, partly by a slight invection of the wall and
partly by a thinning of the ectoderm and mesoglcea so that at
length they may be as deep as half the thickness of the wall,
but they never become like the deep pits of the preceding two
genera. No acontia could be found.

This genus resembles, in some respects, the Paractide, but
it seems to belong to the Bunodactidw, on account of the
sharply circumscribed sphincter muscle, although the latter
is not so much separated from the inner layers of the wall as
usual in that family.

Epigonactis, gen. nov.

Large stout Bunodactide, with numerous large egg-pits on
the upper half of the column, arranged in more or less regular
rows and often crowded on the upper part, below the collar.
These pits, when fully formed, are deeply excavated in the thick
mesoglcea and when containing eggs some of the latter may be
nearly or quite covered and concealed by the tissues growing
around them. In contraction there is a well-developed fold or
collar covered with numerous longitudinal grooves, but with-
out verruce; fosse smooth. Tentacles numerons, large, rather
long, suleated in contraction. Sphincter muscle large, cireum-
scribed. Perfect mesenteries numerous and fertile. Eggs |
large. Wall firm and thick, in sections the layers are very dis-
tinct ; the circular muscular layer shows, in alcohol, as a well-
defined, continuous, brownish yellow layer; mesogloea thick.

Epigonactis fecunda, sp. nov. Figure 35.

Body nearly cylindrical, about as high as broad in contrae-
tion; base muscular and strongly adherent, not much wider
than the column, or even narrower in one example in alcohol.
Tentacles, in the smaller specimen, 72, stout, blunt, some of
them rather long, others contracted so that they are only
little longer than thick; they are retracted in most cases so
that their tips project but little beyond the marginal collar but
are not infolded ; some of them appear to be perforated at tip,
and most are suleated by contraction. They are arranged in two
or three marginal rows. Collar large and prominent, much
thickened, but without marginal tubercles; fosse deep, with a

A. FE. Verrill—New Actinians. 379

nearly smooth surface. Outside of collar and upper parts of
column below it are covered with intricate wrinkles and irregular
vermiculate elevations, which surround and enclose numerous
deep round pits, mostly closely crowded together. In most of
the pits there is a large spherical egg, about 2™™ in diameter ;
most of these project beyond the pits, but some are entirely
concealed in the thickened integument.

The egg-bearing pits are most numerous within 12 to 15™™
of the collar margin, but some occur lower down. They form
a conspicuous zone all around the body, in which the eggs are
so crowded as to be in contact in the alcoholic specimens.
They adhere firmly. Lower down on the column there are
sucker-like pits in vertical rows, decreasing in size downward
and mostly disappearing near the middle. They are obscured
by the wrinkled surface, but become very evident where the
soft wrinkled ectoderm is rubbed off.

In vertical section the wall is moderately thick and very
firm both in base and column, plicated by wrinkles; the layers
are easily visible with the naked eye, both in base and column ;
the circular muscular layer shows as a continuous yellow line.
The mesogloea is variable in thickness, being thickened near
the base and in the basal disk, but especially so at and below
the collar, where the egg-pits occur; in this region it is very
much thickened around and between the pits, but is very thin
under them, so that it is very irregular and much lobulated ;
above the collar it is thin. The sphincter muscle is large,
round or ovate in section, circumscribed and pedicellate. The
mesenteries form about 36 pairs, of which 24 or more are
perfect, but they are irregularly developed in the different
sextants. In one case six fully developed and nearly equal
pairs occur on one side of a pair of directives, while on the
other side those of the fourth and fifth cycles are much smaller
than the rest. The perfect pairs and most of the others bear
large and broad longitudinal muscles, which .are thick and
strongly plicated. All or nearly all the mesenteries bear
large gonads, which are so densely crowded together, in the
specimen dissected, that it is not possible to separate them, or
place them accurately. They contain numerous large ova
similar in size and appearance to those on the exterior of the
body. Color not recorded.

Height, in alcohol, about 2°25 inches (56™™"); diameter 2°5
inches (62™"™).

East of Banquereau, off Nova Scotia, 150 fathoms, Sch.
“Polar Wave,” lot 37, 1878, 1 specimen; off Nova Scotia, 200
fathoms, Capt. John Rowe, lot 35, 1 specimen.

380 A. E. Verrill—New Actinians.

Epigonactis regularis, sp. nov. Figure 36.

Body, as contracted in alcohol, nearly hemispherical, with
the base as wide as any part. Disk and tentacles concealed.
Wall firm and elastic, covered with a yellowish brown ecto-
derm; not much wrinkled, except near base and on collar.
The collar is strong and suleated, and also finely wrinkled
transversely, giving the surface an areolated appearance.
Near the base the wrinkles are transverse and wavy. On the
upper half of the column there are vertical rows of large
ego-pits, some of them circular and cup-shaped, others oval or
ego-shaped ; some of them are bordered by several small con-:
vergent lobes, others with folds, while some are widely open.
They carry no eggs in the type, but are evidently used for
that purpose. The wall is thickened at the collar, but thin
above it; lower down it is of moderate thickness but firm ;
the three layers are very definite and easily visible to the
naked eye, in sections, without'staining ; the sphincter is large,
elliptical in section, well circumscribed. The circular muscle
layer 1s continuous and well-defined.

Tentacles about 90, stout, suleated, rather obtuse, of mod-
erate length, arranged in three or four crowded rows; mouth
large, with two siphonoglyphs, bordered by large lobes.

Mesenteries form about 44 pairs, most of which are perfect
and bear gonads. In a section, near the lower part of the
stomodeeum, 36 perfect pairs were counted, and most of the
remaining eight, which were nearly as wide, may be attached
lower down. Those next to the two pairs of directives, on
each side, were the least developed. Owing to the equality in
the size of the mesenteries no definite hexamerous arrange-
ment could be made out. The longitudinal muscles are very
broad, covering nearly the whole breadth of the mesenteries,
thick and pleated; those of the directives are less thick, but
have a. second inner enlargement, near the siphonoglyphs.
There are no rudimentary or narrow mesenteries between the
larger ones in the gastric region.

Fishing Banks, off Newfoundland, in deep water. It was
brought up on the trawl hooks of a Gloucester, Mass. fishing
vessel, but no record of the precise locality was given.

Rh. G. Leavitt— Cause of Root-Pressure. 381

Arr. XLIV.—A Preliminary Note as to the Cause of Root-
Pressure ; by Ropert G. LEAVITT.

THE following suggestion is made in the hope of breaking
a way toward the proper understanding of the nature of so-_
called root-pressure. The hypothesis here tentatively proposed
seems to coordinate certain of the results obtained by Sachs,
Clark and others, which after several years still stand in appar-
ent need of explanation.

Van *t Hoff’s interpretation of the exact determinations of
osmotic pressures made by Pfeffer, de Vries, Tammann and
others, which interpretation is generally accepted by physicists
and by Pfeffer himself, though still neglected by the majority
of writers on vegetable physiology, is here assumed as the
basis of the discussion on the physical side. Unfamiliarity with
this theory on the part of many botanists may warrant a con-
densed statement of its central principles.

Solutions (the term being applied to the dissolved sub-
stances, exclusive of the solvents) act in all ways, both qualita- »
tively and quantitatively, as gases, exerting pressure against
containing water-, or other, surfaces by an inherent expansive
force. The laws of solutions are identical in form, if “solu-
tion” be substituted for “body of gas,’ with the laws of
Boyle, Charles and Avogadro for gases. Quantitatively con-
sidered, “dissolved substances exert the same pressure, in the
form of osmotic pressure, as they would exert were they
gasified, at the same temperature, without change of volume.”

The distention of a turgid cell is then due to the pressure of
organic matters in the cell sap, unable to escape or able to
escape only with difficulty. The entrance of water is a con-
sequence, not cause, of the osmotic pressure. A_ suflicient
supply of water at hand is, to be sure, a necessary condition
for allowing the osmotic force to act and push the cell-wall
outwards, since the boundaries. of the solvent, equally with
definite membranes, limit the molecular wandering of the
solution.

The movement of water up the stem of the thistle-tube, in
the familiar experiment, is caused by the upward pressure of
the solution against the free surface of the water; the water
and the membrane below both being passive accessories in the
operation. If the stem were crossed by diaphragms _per-
meable by both water and solution, these would not constitute
osmotic membranes in the sense of occasioning osmotic stress
and flow of water; osmose and filtration of water would pro-
ceed through them. If the stem, however, were crossed by a
membrane permeable by water but not by the solution, both
would be stopped, and no exfiltration of the water by action
of the confined osmotic substance is conceivable.

382 R. G. Leavitt—Cause of Root-Pressure.

Osmose being defined as the diffusion of a solution through
a membrane, and water being in no sense a solution, “ endos-
mose of water” is mistaken “phraseology. Some other appro-
priate ter m should be employed, such as infiltration. ‘ Osmotic
pressure” is unmistakable, but. it is well to keep in mind
that pressure arises through failure of the solution to accom-
plish osmose.

Coming now to the plant-body as an osmotic machine, we
may distinguish two categories of osmotic pressures with their
consequent filtrations of water, according as (1) the solutions
are confined within single (living) cells, or (2) the osmotic sub-
stances permeate the whole cellular system from root-tip to
leaf. The first may be termed localized or intracellular
osmotic pressure, the second general internal, or sap-pressure.

The difficulties involved in ascribing root-pressure to local-
ized osmotic action of the root parenchyma are well under-
stood, and need not be dwelt upon. Some of them seem to be
obviated by attributing the phenomenon in question to general
internal osmotic pressure. The root immersed in the wet
soil may be compared to the thistle-tube of the common
experiment. The root-surface or surface tissue will then
answer to the membrane stretched over the mouth of the tube.
Trachese, tracheids, intercellular spaces, and all parts dead or
alive through which sap-solutions percolate, may answer as a
whole to the interior of the thistle-tube. Soil water may cor-
respond to the water in the outer dish and sap with its con-
tents, to water with its salt or sugar inside the thistle tube.
The water in the stem of the latter, rising by reason of
upward osmotic pressure on its free surface, overflows in an
osmotically driven stream. ‘The outflow from the root, when
the stem is cut off, may have a like cause. In the artificial

apparatus the solution washes out with the stream, and the

end of the process ensues. In the case of the living apparatus
stores of insoluble organic matters (as starch) are present, from
which the reduced fund of osmotic substances (as sugars) may
be replenished from time to time by action of enzymes, causing
periodic overflows of the kind described by Sachs. The flows
would naturally decrease in volume as the stock of reserve
materials became gradually exhausted.

The observations of Sachs on the effect of heat in increasing
absorption, and of Olark on pressure in tree-trunks, accord well
with the supposition that both kinds of phenomena are due to
general osmotic sap-pressure.

The agency of osmotic sap-pressure as more or less import-
ant among several possible factors in maintaining or restoring
the continuity of free water between root and leaf during
transpiration, is here suggested. So-called negative pressure
may probably be shown to be compatible with a simultaneous
osmotic sap-pressure of considerable intensity, when the supply
of water at the root falls below the demand.

Harvard University, April 14, 1899,

7
P
&
;

:
-

G. R. Wieland—American Fossil Cycads. 383

Art. XLV.—d4 Study of Some American Fossil Cycads.*
Part Ul. The Female Fructification of Cycadeoidea ; by
G. R. WrELAND. (With Plates VILi-X.)

Introduction.

FEMALE fruits of fossil Cycads from America have not been
hitherto described. European specimens, however, have fur-
nished two nearly perfect examples which have been made the
objects of careful research and discussion during the past thirty
years. In fact, the study of Bennettites Gibsonianus by
Carruthers,’ and Solms-Laubach,®* together with Lignier’s”
investigation of Bennettites Morierie, may well be considered
one of the triumphs of modern paleobotany. As a result, the
structure of the cycadeoidean seed-bearing organs is nearly as
well known as that of the existing cycads.

The subject of female fructification, as far as treated before
the announcement of the discovery of the corresponding male
flower in Part I of these contributions, has been amply
reviewed by Seward. Hence it is only necessary to state
briefly that the object of this preliminary paper is to deal with
several quite perfect cycadeoidean fruits as observed in cer-
tain trunks from the Jurassic of the Black Hills. These, it
will be seen, are in all essential characters similar to the
English, French, and Italian examples of Cycadeoidea, of which
the best known are Cycadevidea ( Williamsonia) gigas, Car-
ruthers, and C. (Bennettites) Gibsonianus, Carr., both from
the lower Greensand of the Isle of Wight, C. (Lennettites and

Willamsonia) Moriérie, Saporta, from the Oxfordian of Nor-
mandy, and (. etrusca, Capellini and Solms-Laubach, from the
Etruscan Necropolis at Marzabotto, the original locality of
which is unknown. The American forms are so analogous to
the several foregoing species as to preclude more than specific
difference, and thus add one more argument in favor of the
approximately synchronous existence of these highly special-
ized forms on both continents. Whether or not final study will
permit their retention in a single genus is, however, a doubtful
question that need not now be taken up; nor can the subject of
the bracts and peduncles be treated in the present paper,
although recent studies by Scott* make it quite certain that
their investigation will furnish valuable data.

* Part I of these studies appeared in this Journal for March of this year, and

Part II in the April number following. The numbers refer to the list of papers
given at the end of Part I.

384 © G. R. Wieland— American Fossil Cycads.

The Fruit of Cycadeoidea Wielandi, Ward (MS.). 2

General Featwres.—Cycadean trunk No. 77 of the Yale
collection, illustrated on Plate VIII, figure 1, is the female type

of the foregoing species. Itisa finely preser ved example, bear-

ing laterally, in addition to ten apparently fresh scars left by
the shedding of ripened fruits, forty more or less complete
seed-covered fruits, which were mature or nearly so when
fossilization began. The shape of these fruits is ovoid, and
their average size is from 3°5 to 4™ in length by 2 to 2: 5am in
diameter. In both size and form, they resemble the fruit of
Cycadeoidea (Bennettites) Gibsonianus, Carr., though the
seeds are less elongate in the latter species.

The embryos of the seeds of this trunk prove to be the only
important structures not present in unusual clearness of detail.
Moreover, this was the first seed-bearing cycad to be described
as such from American formations, as well as the one upon
which these studies were begun. Hence it is made the first
type of female fructification in American. cyeads, although

other more perfect supplementary material has since been -

studied.

A close examination of the specimens represented in the
plate affords interesting facts concerning the fruit-bearing
habit of Cycadeoidea. In their various stages of growth,
maturity, and dehiscence, the fruits are borne on short “pedun-
cles approximately of the same length, thus indicating that the
mature fruit was not protruded further from the trunk by
subsequent growth. Nor does there seem to be any order of
maturity from the base to summit of the trunk in either
direction.

In several other related trunks most of the fruits have been
shed, being represented by scars only; while a few floral axes
remain undeveloped or in an atrophied condition. Still other
trunks of this and associated species are barren or bear embry-
onic fruits. The latter condition is of considerable importance,
as it will permit a developmental study of the organs. The
only general conclusion to be drawn at present, however, is
that there was no sharply limited period when these eycads
matured their flowers. Apparently there was a progressive
maturity of fruits through an entire season, which was followed
by several seasons of unfruitfulness.

From the characteristic form and seulpturing of the scars
representing the position of the matured fruits, it would seem
that the latter were shed by the parent plant with their parts
intact and often with few or no accompanying bracts.

Following fructification and the shedding of ripened fruit,
‘there was a closing up of bracts about the receptacle, with

J

G. R. Wieland—American Fossil Cycads. 385

more or less obliteration. The appearance of the fruits as
apparently mature on the trunk, and of the scars mentioned

above, are fairly well represented i in figures 1 and 2, Plate VIII.
In all cases where the seeds are exposed the tips of the bracts
are necessarily more or less worn or broken away.

The lateral, or proliferous, inflorescence here exhibited is
noteworthy. In this respect these specialized culminant fossil
types vary from all the higher existing forms which are sym-
podial, and partially agree ‘with the genus Cycas. The latter
is the most primitive living form, and has its peduncles dis-
tributed between the petioles for some distance below the apex
of the trunk. Among the various species represented in the
Yale collection there are none showing the sympodial type and

but few flower-bearing examples which have no flowers on the

lower half of the trunk, though the male trunks incline to bear
the most of their flowers near the summit.

Special Organs.—As stated above, these fruits are ovoid
bodies borne on short peduncles. The latter take their origin
on the trunk after the manner of the petioles and terminate in
an elongated receptacle. This bears laterally an involucre of
slender enveloping bracts, and terminates in a “ parenchyma-
tous cushion,” as termed by Carruthers.* The face of this
cushion, or secondary receptacle, is flatly convex and bears two
series of densely set organs, much modified according to their
central or more peripheral position. The first of these are the
seed-bearing stems, or essential organs, while the second are
merely interstitial and quite analogous to chaff. They are
termed “interseminal scales” by Lignier.

The disposition of these parts is illustrated in the longitudi-
nal sections, figures 10 and 12, Plate LX, and in the correspond-
ing transverse sections, figures 9 and 11 of the same plate, as
well as in the several transverse sections on Plate X.

In this arrangement of densely set seed-bearing stems and
surrounding scales borne on a fleshy bract-enveloped receptacle,
there is a certain simulation of the flowers of Composite.

Histological Structure.

The Parenchymatous Cushion.—This is composed of soft
parenchyma centrally, but becomes more dense peripherally.
Though no well-marked bundles have been noted, sparse and
irregularly branching strands of scalariform cells apparently
supply the seed stems. As there is usually a cleavage line, or
zone of imperfect preservation, just at their origin, this could
not be readily determined. The strands may scatter and end
in the cushion. They are illustrated by Carruthers in his
description of Bennettites Gibsonianus.

386 G. R. Wieland—Amertcan Fossil Cycads.

Seed Stems.—These are very closely set on the cushion, and
as it is difficult to separate their elements from those of the
interseminal scales in the longitudinal sections, they may be
best studied in transverse sections. Such a section from the
middle of the fruit is shown in figures 14 and 16, Place X,
where the stem is seen to consist of a central vascular bundle
enclosed in an endodermis which is surrounded by a broad
woody outer layer, or xylem zone, the “ enveloppe tubuleuse ”
of Lignier. ~

Each seed stem is uniformly surrounded by from 5 to 6
interseminal scales. A series of transverse sections from the
base to the summit shows no variation from this general
arrangement, although there is a slow diminution in the
diameter of the stems, accompanied by marked changes in the
cell conformation of the xylem zone. In this and other
species, there is often at the base a strong compression of
the surrounding scales, while the xylem areas are more
or less polygonal and composed of cells qnite regularly sub-
rhombic in section. Of these, from 30 to 40 may be counted
radially, as shown in figure 4, Plate VIII. Towards the
summit there is a constant increase in the size and regularity
of these cells, which finally become quite round, accompanied
by a decrease in the width of the xylem band, until at the
base of the seeds it consists of but two layers of cells surround-
ing the central bundle, which remains constant in character.

This bundle is well illustrated in the enlarged view given in
figure 15, Plate X. Centrally there are from 3 to 10 small
woody elements. These are enveloped in soft parenchyma,
not preserved in all specimens, and the bundle terminates in a
well-marked endodermis. Such concentric bundles recall the
somewhat similar structures seen in many ferns. The
atrophied peduncles, or those lacking a well-marked central
bundle, described by Lignier in C. (Bennettetes) Morierce, have
not been observed thus far. The interesting case of bifid
seed stems, or those with two central bundles within the same
tubular, or xylem zone, however, rarely repeats itself in the
. present specimens.

It may be mentioned that while the figure of the transverse
section of the seed stems, ‘“ vascular cords,” given by Car-
ruthers is a faithful one, its being cut too near the base, where
there is much compression of the several elements, prevented a
true conception of their significance. Only at some distance
above the base does the clear differentiation of parts seen in
figure 14, Plate X, appear. Judging from the transverse
section by Carruthers, just mentioned, and the additional one
given by Solms-Laubach, the tubular layer, or xylem zone, of
C. (Bennettites) Gibsonianus was composed of more angular
cells than in the case of the present specimen.

3 eS ae ar

= Sow

SSS ye et ee Fe

Te

G. R. Wieland—American Fossil Cycads. 387

Lignier gives a section of a seed stem from near a seed base
(peduncle, figure 27, Plate 1I),” in which the tracheidal cord is
followed by fundamental tissue, and this by a layer of cells
(assize colorée), which is in turn surrounded by a single row of
cells laterally dovetailing. This alteration at the seed bases
has not been observed as yet by the writer, and no doubt must
stand as one of the specific variations between the foregoing
species.

Interseminal Scales.—Five or six of these commonly sur-
round each seed stem, and rise beyond the seeds, their tips
expanding to form the rounded summit of the fruit. Figure
13, Plate X, displays a transverse section of one of these
organs cut near its middle height. In the case of cycad
No. 77, figure 14, Plate X, the parenchyma surrounding
the central vascular strand is not preserved; but the central
strand, the large subepidermal cells (much less continu-
ous in this species than in C. Morierie), and the epidermis, are |
so uniformly present as to differentiate these tissues as abso-
Iutely as any staining. From the base toward the summit, as
the seed stems decrease in diameter, the scales increase in size
by the enlargement of the cross area of the parenchyma cells,
which are usually rounded in section and quite elongate. Be-
yond the seeds, the cells composing the expanded summits are
entirely lignified. Intheseed zone, the scales are also more or less
hignified and often very thin and flat as they pass between the
seeds, where it is sometimes difficult to differentiate them from
the seed coats. Exterior to the outer seed stems there is a
flattened overlapping series of scales, forming a kind of lateral
wall of the fruit several scales in depth. The distal portions
of these modified scales are elongate, tetrangular, and lignified,
as they appear on the fruit surface.

It is difficult to detect the several elements of the scales in
longitudinal section. That some of the xylem cells are spir-
ally marked is certain, and that there is an occasional scalari-
form marking is also certain. The woody elements of these
scales regularly take their origin below that of the cells of the
xylem zone of the seed stems, and in figure 5, Plate VIII, are
seen extending down into the parenchymatous cushion as small
projecting brushlike groups.

It need scarcely be pointed out that there is an analogy be-
tween these scales and the central bundle of the seed stems, the
variation being in a uniformly large central strand, the addition
of a broken subepidermal layer with apical sclerotization, and
the presence of occasional ducts in the case of the scales.

The Sceds.—The seeds thus far observed in the present and
other species show much variety of preservation. Their study
is therefore a difficult one, requiring time, especially for the

Am. Jour. Sc1.—Fourts Series, Vou. VII, No. 41.—May, 1899.

26

388 G. R. Wieland—Americun Fossil Cycads.

purpose of comparing a large number of the same and differ-
ent species, and determining how far the structures supple-
ment each other. Hence no detailed account of seed structure
will be now attempted, although the appended figures, which
in reality are but little diagrammatic, are introduced.

1 2

--- a4

ae
ees:

Al iee

---€

BUS!

----9

FIGURE 1.— Cycadeoidea (Bennettites) Moriérie, Saporta 2; longitudinal section of
seed. (After Lignier.)

a, micropylar tube; 6, prismatic layer; c, pulpy tissue; d, corpuscular mass;
e, interseminal scales; 7, embryo space; g, remains of nucellus; h, chalaza; 2,
tubular layer (= xylem zone); #, micropylar canal; 7, nucellar beak; m, pollen
chamber; n, fibrous stratum; 0, basal expansion of 2; p, parenchyma of seed
stem.

FIGURE 2.—C. Wielandi, Ward 2; longitudinal section of seed. x 12

a, micropylar tube; 0, interseminal scale; c, the same; d, cotyledons; e,
fibrous layer ending below the chalaza in an expanded base of scalariform tissue ;
f, a structureless zone; g, stringy remains of cells lignified and slightly expanded
below; h, chalaza; 7, woody scalariform tissue; #, prismatic outer layer of seed
ending in xylem zone; /, parenchyma.

General Considerations.

Owing to the large number of male and female examples of
Cycadeoidea in the Yale collection, their proper correlation will
require considerable time. Nevertheless, several facts of im-

G. Rk. Wieland—American Fossil Cycads. 389

portance present themselves for mention. A number of species
bear fruits of the type represented in the well-known figure of
C. etrusca given by Capellini and Solms-Laubach (Plate IV).*°
These heart-shaped fruits have very closely set, rodlike seeds
arranged prismatically on a short, fleshy axis, traversed by nu-
merous gum ducts and usually devoid of the regularly arranged
seed stems and interseminal scales characterizing C. (Bennett-
ites) Gibsonianus, C. Morierie, and C. Wrelandi. Should
_ there prove to be no intervening forms and no connection
_ during developmental stages, this must be finally regarded as
a strong generic difference. One of the Yale specimens not

yet fully studied, but nearly related to C. etrusca, as far as
may be judged from general features, displays seeds with well-
_ marked cotyledons, as represented in figure 7, Plate IX.
_ Another closely related trunk, of which the female form also
is known, shows well-preserved male cones of essentially the’
same character as those described in Part I of these contribu-
tions. Pollen grains, both distended and dessicated, and repre-
sented in figures 9-16, are present in rare preservation. They
are so like the dissociated pollen figured by Capellini and
- Solms-Laubach” in their description of C. etrusca that figures
17-20 in the text are introduced for comparison, as well as
pollen from the living cycad Ceratozamia, figures 3-8. That
these fossil bodies represent characteristic cycad pollen is be-
yond all question. Sections of the cycadeoidean cone from
which the pollen of figures 9-16 was derived are given in
figures 17 and 18, Plate X.

FIGURES 3-8.—Ceratozamia longifolia, Miquel; pollen grains. (After Juranyi.)
x 610.
3, mature pollen in dry condition; 4, 5, the same swollen in water; 6-8, the
respective transverse sections, or end views,
FIGURES 9-16.— Cycadeoidea, sp., Black Hills; pollen grains. x 250.
9, 10, 13-15, more or less dessicated grains; 11, 12, 16, normal forms.
FIGURES 17-20.-—— Cycadeotdea etrusca, Capellini and Solms-Laubach; isolated and
dried pollen grains. (After Capellini and Solms-Laubach).

390) G. R. Wieland—American Fossil Cycads.

But one interpretation of these facts is possible—Cycade-
oidea etrusca represents a fertilized dicecious plant, the male
flower of which has now been definitely recognized in an
allied species. The so-called “antheriferous tissue” noted by
Capellini and Solms-Laubach hence remains an unexplained
structure, possibly of much significance. Should this tissue
represent a developmental stage instead of permanent structure,
as certain sections made by the writer suggest, it nevertheless
merits the closest examination, and may have some connection
with a “corona” such as that of Williamsonia gigas.

Another close resemblance between the American and
European species having prismatic seeds is the characteristic
surface sculpturing. Figure 2, Plate VIII, is introduced here for
comparison with a Bennettitean specimen illustrated by Seward
(Fossil Piants of the Wealden, Plate XV, figures 5 and 6).*

" Yale Museum, New Haven, Conn., April 24, 1899.

EXPLANATION OF PLATES.

PLATE VIII.

FIGURE 1.—Cycadeoidea Wielandi, WardQ; type specimen. x+H. ©

FIGURE 2,—Cycadeoidea (sp.) 9; surface sculpturing of a fruit. x10.

FIGURE 3.—Cycadeoidea Wielandi, Ward 9; portion of a trunk, showing scars
left by the shedding of fruits. x4.

FIGURE 4.—Cycadeoidea turrita, Ward Q; transverse section of a seed stem
(from near the parenchymatous cushion) surrounded by an unusual
number of interseminal scales. x30. . .

FIGURE 5.—Cycadeoidea Wielandi, Ward 9; longitudinal section ‘through seed
stems. showing their origin on the parenchymatous cushion. x 24.

FIGURE 6,—Same fruit as figure 5; transverse section of seed stems near
periphery of fruit. x30. (Compare with figure 14.)

PLATE IX.

FigurRE 7.— Cycadeoidea (sp.) 2; near ly transverse section through the periphery
of a fruit, cutting 5 seeds, in several of which the cotyledons may
belseem: 2. x 16.

FIGURE 8.—Cycadeoidea Wielandi, Ward 9;. longitudinal section through a fruit
near the summit, cutting superimposed seeds. x 12.

FIGURE 9.—Cycadeoidea Wielandi, Ward 9; type specimen; transverse section
through fruit; showing seeds, seed stems, and interseminal scales.
x 4.

FIGURE 10.—Same fruit as preceding; longitudinal section through seed stems and
parenchymatous cushion. x4.

FIGURE 11.—Cycadeoidea turrita, Ward 9; transverse section through fruit,
cutting a single seed obliquely. x 25.

FIGURE 12,—Same fruit as preceding; longitudinal section. x 4.

a, seed; 6, densely set interseminal scales and aborted seed

stems (?); c¢, the parenchymatous cushion; d, nyo nies
é, line of dehiscence (?).

* Tn the case of genericand qectia. names, the writer has followed Prof. a
F, Ward of the U- ie _National Museum, as the recognized authority on American
cycads. Fortunately for scientific ‘uniformity, Prof. Ward has described practi-
cally all the cycads discovered in this country.

G. R. Wreland—American Fossil Cycads. Bot

PLATE X.

FIGURE 13.—Cycadeoidea Wielandi, Ward 9; transverse section of an inter-
seminal scale. x 185,

a, xylem zone of a bordering seed stem; 0, epidermis of the
interseminal scale; c, xylem; d, parenchyma; e, heavy and much
broken subepidermal layer. :

Note.— Where two scales border there is a double layer of 0.

eee 14.— Cycadeoidea Wielandi, Ward 9; type specimen; transverse sec-
tion through seed stem and surrounding interseminal scales, no
parenchyma of either of these organs being preserved. x 22.

FigTRE 15.—Same fruit as figure 13; transverse section through the central
bundle of aseed stem. x 185.

a, xylem zone; 6, endodermis; ¢, parenchyma; d, xylem; ¢, cel-
lular interspace.

FicguRE 16.—Same fruit as preceding; transverse section of seed stems and
interseminal scales, all elements being preserved. x 22.

FIGURE 17.—Cycadeoidea (sp.) 6; transverse section through the pendent sori
of asporophyll. x 18.

a, epidermis of prismatic cells which entirely invests each sorus,
the latter consisting of two parallel rows of sporangia; b, one of
the sporangia containing pollen grains as illustrated in text, figures
7-14, page 389.

FIGURE 18. —Same as preceding, but showing a ee area. The arrow points
towards the floral axis.. x3.

392 Scientific Intelligence.

SCIENTIFIC INTEHEDIGCEWN Cee

I. CHEMISTRY AND PHYSICS.

1. On Low Temperature Haperiments.—In consequence of the
easy production of liquid oxygen and nitrogen in recent times,
experiments at low temperatures have become common. HEMPEL
has called attention to the fact, however, that with much simpler
means than these, temperatures suflicing for the greater part of
the ordinary reactions can be secured. Thus, for example, liquid
carbon dioxide, now commercially very cheap, may readily be
employed to show that potassium and bromine no longer react on
each other at low temperatures. If the bromine be first placed
in a mixture of solid carbon dioxide and ether and thoroughly
cooled, and then a piece of potassium, cooled in a similar bath,
be introduced into it in a small spoon, no combination takes
place; though it speedily sets in as the temperature of the mix-
ture rises. Very considerable improvements, however, are neces-
sary in the methods of manipulation. The double-walled and
exhausted globe of Dewar, therefore, has been of great service.
But these Dewar globes are costly and fragile when they are of
large size; and hence are not available for laboratory experi-
ments on any large scale. The author therefore has experimented
to determine to what extent a sufficient isolation can be secured
in other ways. Test tubes 4° in diameter, approximately of the
same size as the Dewar tubes, were placed in beakers 13™ in
diameter, the space between being packed with the material to be
tested. Equal quantities of solid carbon dioxide and ether were
placed in each of these tubes; so that at the beginning of the
experiment all the vessels contained equal quantities of material
at —79°. From time to time the temperature in the several tubes
was determined by an electric pyrometer. In 5 minutes the
tube packed with cotton had a temperature of —76°, that with

wool —77°, that with eider-down —78°, and that with a well

exhausted Dewar tube —77°. After 58 minutes these values
were successively —56°, —64°, —67°, and —54°; thus showing
that well dried wool and eider-down are very good insulators,
being exceeded only by the best Dewar tubes. For preparing
the solid carbon dioxide the author prefers a paper funnel with a
wide short neck, to the upper edge of which is fastened a cylin-
drical bag of linen. On clasping the upper portion of this bag
about the valve on the cylinder containing the liquid gas, and

opening this valve, the solidified gas collects in the bag and may ~

be readily shaken into the funnel. About 270-300 oms. of the
solid are obtained from a kilogram of liquid; and since this liquid
costs about 15 cents a kilogram, the solid would cost not far
from fifty cents a kilogram. No advantage in yield was found to
result on cooling the liquid before expansion. The lowest tem-
perature is obtained when the solid carbon dioxide is mixed with

ew eee ee ee ee ee a a?

Chemistry and Physics. 393
ether to a stiff magma.— Ber. Berl..Chem. Ges., xxxi, 2993-7,
December, 1898. G F. Bz
2. On Reductions in Presence of Palladium.—Although pal-
ladium-hydrogen has been used already as a reducing agent by
Graham and by Saytzeff, yet now by a more extended study
Zevinsky finds that it may be employed with great success in
the reduction of bromides and iodides of cyclic alcohols, a class
of compounds which hitherto it has been impossible to reduce by
any of the ordinary methods. The form in which it is used is
that of a palladium-zinc couple, analogous to the copper-zinc
couple of Gladstone and Tribe. To form it, zine filings are well
washed with alcohol, then treated with sulphuric acid until an
active evolution of gas takes place, and finally washed with
water. On pouring upon them a one or two per cent solution of
palladium chloride previously acidified with hydrochloric acid, a
thin strongly adhering layer of palladium is deposited upon the
zinc. The couple thus formed is washed with alcohol and dried.
To effect the reduction, which takes place at the ordinary temper-
ature and up to 100°, a flask provided with a reflux-condenser and
dropping funnel is one-third filled with the palladium-zine couple,
this couple is partly covered with methyl or ethyl! alcohol and hy-
drochloric acid saturated at 0° isrunin drop by drop. At first the
hydrogen is absorbed by the palladium; and as soon as bubbles
of gas are evolved, the bromide or iodide to be reduced is added
gradually, alternately with the acid. The action is controlled by
regulating the quantities of material added. ‘The evolved gases
are passed through alcohol to retain any volatile hydrocarbon
which may be carried over. The palladium-zinc in the flask, after
washing with acid and with alcohol, is ready for further use.—
Ber, Berl. Chem. Ges., xxxi, 3203-5, January, 1899. «GF. B.
3. On the Properties of Metallic Calcium.—Further results
have been communicated by Motssan on the properties of calcium.
Prepared by the method already described,* the metal separates
from fused sodium in hexagonal crystals which have a specific
gravity of 1°85 and melt at 760° in vacuo. On solidifying, the
metal is somewhat brittle, is less malleable than potassium or
sodium and shows a crystalline fracture. When free from nitride,
it is silver-white in color and has a brilliant surface. Heated to
redness in a current of hydrogen, a crystalline hydride CaH, is
formed. When pure, calcium is not acted on at ordinary tem-
peratures by chlorine; though at 400° the action is decided. But
if the metal contains nitride, chlorine: attacks it at the ordinary
temperature. At 300° it ignites and burns brilliantly in oxygen.
Gently warmed in air, it burns with brilliant scintillations. It
combines with sulphur at 400° with incandescence. At a red

heat it unites actively with lamp-black, yielding CaC,. Calcium

gives somewhat brittle alloys with magnesium, zine and nickel.
With tin the alloy contains 3°82 per cent of calcium, has a density
of 6°70, and slowly decomposes water. <A crystalline amalgam is

* This Journal, IV, vi, 428, November, 1898.

394 Scientific Intelligence.

formed with mercury which may be distilled in hydrogen at 400°,

but which forms nitride when heated in nitrogen. Heated to red-
ness with potassium or sodium chloride, calcium sets the metal
free. Water acts on calcium only very slowly, evolving hydro-
gen. At a red heat, it combines explosively with phosphoric
oxide, and decomposes silica, yielding calcium silicide and some
silicon. At 600° it reduces boric oxide to boron. Nitric acid
when strong acts on calcium very feebly, though when diluted the
action is rapid. Fuming sulphuric acid is reduced by the metal.
Heated to dull redness in ammonia, a mixture of hydride and
nitride is produced. In liquefied ammonia at —40°, calcium
ammonia is formed, a reddish brown solid.—C. #., cxxvii, 584—
590, October, 1898. G. F. B.
— 4. On New Bases from Strychnine.—It has been noticed by
Ta¥FeEL that when a current of five amperes at a potential of three
and a half volts is passed through a solution of 30 grams of
strychnine in 180 grams of concentrated sulphuric acid diluted
with 120 grams of water, two poisonous bases are produced which
he calls tetrahydrostrychnine and strychnidine. After passing
the current for two hours, the liquid, diluted with four volumes
of water, is nearly neutralized with barium carbonate, filtered,
heated to boiling and treated with barium hydrate in slight
excess. The strychnidine is contained in the precipitate while
the filtrate contains the tetrahydrostrychnine. Strychnidine crys-
tallizes from alcohol in stellate groups of colorless needles,
fusing in vacuo at 252° and having the structural formula

CvER IOS a " It boils at 290°-295° under 14™ pressure, and

may be distilled undecomposed, the distillate solidifyimg to a crys-
talline mass. It is sparingly soluble in water, but dissolves more
freely in alcohol, benzene and chloroform. A 6:4 per cent chloro-
form solution has a specific rotatory power [a], == — 8°28° at 20°.
Dissolved in sulphuric acid, it gives color with oxidizing agents.

Its aqueous solutions with excess of acid become intensely red —

with potassium dichromate, hydrogen peroxide, ferric chloride
and sodium nitrite. Tetrahydrostrychnine C,,H,,NO(NH).CH,.
OH crystallizes from alcohol in colorless prisms, containing a
molecule of the solvent,-which it loses at 100°. It dissolves in
280 parts of water at 20°, the solution being strongly alkaline.
It resembles strychnidine in its action with sulphuric acid and

oxidizing agents, but gives a wine-red solution when its hydro-.

chloric acid solution is treated with chromic acid or ferric
chloride. Strychnoline, obtained by reducing dioxystrychnine in
amyl alcohol with sodium, and dihydrostrychnoline, produced by
passing a current through a solution of deoxystrychnine in sul-
phuric acid, are also described.— Liebig’s Annalen, ccci, 285-348,
August, 1898. G. F. B.

5. A History of Physics in its Elementary Branches including
the Evolution of Physical Laboratories ; by Fior1an Cazsort,

q
‘
a
Ey
al
=
i

of

Sryes
ya ae ey

Chemistry and Physics. 395

Ph.D. 8vo, pp. villi, 322. New York, 1899 (The Macmillan
Company).—In this book the author has attempted a somewhat
condensed review of the progress in physical science from the
earliest times to the present. In the preface he quotes Ostwald’s
statement to the effect that a defect attaching to present scien-
tific education is the absence of the historical sense and the want
of knowledge of the great researches upon which the edifice of
science rests. It seems to be fully recognized that the historical
development of a subject is an essential aid to the proper under-
standing of it. it may be a question, however, whether this his-
torical treatment should not be given in connection with the subject
discussed rather than have a separate course devoted to it. In this
book the history of physics is divided into time epochs, 1. e., that
of the Greeks, the Romans, the Arabs, the Middle Ages, the
Renaissance, the Seventeenth Century, the EKighteenth Century
and the Nineteenth Century ; closing with a chapter on the Evo-
lution of Physical Laboratories. This method seems to necessi-
tate considerable repetition, light forexample being considered
six times. Possibly a physical classification would entail less
repetition and be handier for reference. A wide range of refer-
ences has been made use of and these are given in foot-notes.
The book will undoubtedly be of use to students and teachers of
physics, for whom it was intended. .

6. The Electrolytic Interrupter.—Much interest has been ex-
cited by Dr. A. Wehnelt’s new interrupter for induction coils,
which consists of a lead plate, the cathode, and a platinum point,
the anode, immersed in dilute sulphuric acid and connected with
an electrical source of comparatively high voltage. At a recent
meeting of the Physical Society in London, March 10, Mr. A. A.
CaMPBELL Swinton described his experiments with this inter-
rupter. It was found that the minimum working voltage was
twenty-five volts. The spark-pitch varied with the length of the
platinum terminal wire. The wire must be immersed in the.
dilute acid before the circuit is closed, otherwise the action is not
obtained. The interrupter becomes fatigued in a quarter of an
hour and then the platinum becomes red-hot. The interrupter
would be very effective for X-ray work if it did not heat the
Crookes tubes. Mr. Swinton suggested the use of the interrupter
for the production of Hertz waves. Professor Oliver J. Lodge
thought that it would not be effective for this purpose since the
air would not quickly regain a non-electric condition. The
interrupter will work with an alternating current but with half
the efficiency, apparently allowing only one-half the alternations
to pass. A well-known arrangement, however, of circuits might
give a greater efficiency.— Wature, March 16. a

7. Measurement of slow electrical oscillations.—W. Kone dis-
cusses previous methods of. receiving the sparks from an induction
coil on lamp-blacked rotating discs, and describes an efficient
method of measure, which consists of a tuning fork provided
with a stylus which serves both to register the time on a blackened

396 Scientific Intelligence.

disc connected with a pendulum and to serve as a terminal to —
deliver the spark to the moving disc. With this apparatus he
finds that he can follow changes of potential under twenty volts. —
He finds that the electrostatic capacity of the bobbin of the induc-
tion coil is of the order 161071? farad. The working of the
iron core of the coil diminished with increasing number of vibra-
tions.— Wied. Ann., No. 3, 1899, pp. 535-562. Fae

8. Electric waves on wires.—W. D. Cooter has succeeded
in photographing electric waves on wires by using a vigorous
exciter—Blondlot, or Tesla transformer. The photographic lens
was placed above the wires, which were horizontal. The nodes
and ventral segments are clearly distinguished in the photographs.

= Wied. Ann., No. 3, 1899, pp. 578-591. 5. irs

9. Absorption of the X-ra ys by air.—Since very short ultra-
violet waves of light are absorbed by air, it is an interesting
question to consider the absorption of the X-rays by air. Some
recent experiments by Prof. Trowbridge and Mr. J. EK. Burbank
in the Jefferson Physical Laboratory instituted to study this
question show that the difference of absorption of a column of air
thirty inches in length at atmospheric pressure and a column of
air of the same length at four millimeters pressure cannot be dis-
tinguished by the use of a fluoroscope. With cathode rays the
difference is as thir ty-three to one. That is, the absorption is
thirty-three times as much in air at atmospheric pressure as at
four millimeters pressure. The electrostatic effects in the ex-
hausted tube thirty inches long are very beautiful and cease
when the concave mirror of the X-ray tube becomes the positive
terminal. BSL Me

10. Die Optischen Instrumente der Firme R. Fuess-; deren
Beschreibung, Justierung und Anwendung von C. LEIss; pp. Xiv,
397, with 233 text figures and three photo-plates. Leipzig, 1899
(Wm. Engelmann).—Very important advances have been made
within recent years in the methods and apparatus applicable to
crystallographic and optical investigations. The present volume,
giving an account of the instruments furnished by the firm of R.
Fuess (Berlin), is very thorough and complete and will be of.
great value to all working in this field. There are described here,
for example, a number of forms of spectrometers and various
types of apparatus to be used with them, as, for example, the new
spectral arrangement of Wiilfing, giving homogeneous light of
different wave ‘lengths. The reflectometers, spectrographiec « appa-
ratus, goniometers, microscopes, are also fully and minutely
treated of. The chapter giving an account of the different forms
of apparatus for projections with electrical lamps is particularly

interesting and timely. The volume is not simply a catalogue of - q

instr uments, but gives full descriptions of the method of use applic-

able in each case, with references to the original papers. The

worker is thus told not only what instruments are to be had, but

when he is supplied with them, how they are to be adjusted and
used in practical investigations.

Geology and Natural History. 397

II. GroLtocy anp NaAtuRAL HISTorRY. .

1. Note on a Bridger Eocene Carnivore ; by O. C: Marsu.*—
_ A reéxamination of the three species of the genus Limnocyon
- described by me in 1872 (this Journal, vol. iv, pp. 122, 203, and
- 204), leads to the conclusion that the type is in all probability to
be referred to the genus Sinopa, Leidy, from the same horizon.
While this statement applies to Limnocyon verus, the species
first described, yet a second species of the genus, Limnocyon
riparius, described by me in the paper cited, is undoubtedly dis-
tinct, and cannot be placed in any known genus of Kocene
Carnivora. Since the type species of the genus, however, is thus
shown to belong to Sinopa, the generic name Limnocyon must be
abandoned and another substituted. I therefore propose the -
name Zelmatocyon, basing the genus on the remains referred to
Limnocyon riparius.
The generic characters may be stated as follows:
Dentition: I, C;, Pm;z, M,; two lower molars subequal in
size and tuberenlo-sectorial in pattern ; internal cusps of these
teeth considerably reduced, as in Sinopa ; jaw very straight on
inferior border, not unusually deep, but relatively thick and heavy,
symphyseal portion abrupt and robust; first superior molar with
external cusps well separated, and not closely approximated as in
Sinopa.

This genus thus differs from the contemporary Viverravus, in
the unreduced condition of the second lower molar as well as in
the character of the jaw; from Vulpavus, in having two instead
of three lower molars; and from Sinopa as already indicated.

2. The Age of the Franklin White Limestone of Sussex County,
New Jersey ; by Joun Exior Wotrr and Atrrep HutsE Brooks,
Eighteenth Anuual Report of the U. 8. Geological Survey, 1896-
97, Part II, pp. 425-457, 1898.—Messrs. Wolff and Brooks have
made a careful study of the white limestone of Sussex County,
which, with the blue limestone of the same region, has been the
subject of many papers of those attempting to determine the age
of the limestones. Mr. Nason, in 1890, gave at the time what
seemed to be conclusive evidence of the Cambrian age of the
white limestone, attributing its crystalline condition to the intru-
sion of igneous rocks, and also maintained that the white and
blue limestones passed one into the other. The present writers
have reached the conclusion that the white limestones are pre-Cam-
brian in age, and that the quartzite and blue limestone are Cam-
brian. Their conclusion regarding the relationship of these is as
follows: ‘‘'That the white limestone was deformed and meta-
morphosed to its present condition and partly eroded before the
basal member of the Cambrian series was laid down; that the

PE RE ee | Ge oe ih

*This note was prepared by Dr. J. L. Wortman, who first called Professor
Marsh’s attention to the points of synonymy as here given. The matter was left
incomplete at the time of Professor Marsh’s death.—Ep.

i .

ane

398 Scientific Lntelligence.

deformation which the area has suffered since the deposition of —

the Cambrian, which has manifested itself in folding and fault-
ing, has been but slight compared with the pre-Cambrian defor-
mation ; and that the Cambrian rocks overlie the white limestone,
as well as the gneisses, unconformably. As a result of our obser-
vations, we are compelled to adopt the pre-Cambrian age of the
Franklin white limestone.” H. S. W.
3. West Virginia. Geological and Economic Survey.—Pro-
fessor I. C. White, the State Geologist, has issued a map of West
Virginia, compiled by Russell L. Morris, C.E., on the scale of 10
miles to the inch, on which are given the location of oil pools,
natural gas wells, and the approximate distribution of the New

‘River, Alleghany or Kanawha River, and Pittsburg coal areas. |

The map is not complete, but is issued as a preliminary one only.
H. 8. W.

4, Alabama Geological Survey. Iron-Making in Alabama.
Second edition ; by Witt1am Batre PaILytes, pp. 1-380, 1898.—
This is a second edition of the report published in 1896, contain-
ing new statistics regarding the products of the State. Several
new chapters have been added, among them additional informa-
tion as to the coal industry, compiled from the report by James
D. Hillhouse, State mine inspector. H. S. W.

5. The Development of Lytoceras and Phylloceras, by J aMEs
PERRIN Situ, Pro. Cal. Acad. Sci., 3d Ser., Geol., vol. i, pp.
129-160, plates xvi-xx, 1898.—The two genera Lytoceras and
Phylloceras are chiefly interesting ‘“‘ because they are simple,
unspecialized genera, long-lived, little changed, and yet with the
power of giving off other variable branches. They are the two
longest-lived genera of ammonites, ranging from about the end
of the Trias to the Upper Cretaceous, at least seven millions of
years by a conservative estimate.” The author has prepared
beautiful plates illustrating the sutures of the protoconch and
following stages of growth; and concludes from his studies that
‘“‘these two genera come from a common origin, and follow the
same paths up to Vannites, where they part company, each going
through a stage corresponding to that genus, but to different
species under it; both go through Monophyllites stages, but here
again analogous to different groups and even different subgenera.
There the resemblance ceases, and they develop into different
families probably by the middle of the Trias, for in the upper
division of that formation Wegaphyllites and Monophyllites are
sharply distinguished from each other. In the life-history of
these two genera we have a rare opportunity of observing accel-
eration of development, and divergence of two nearly related
stocks, whose history may be traced from the Paleozoic to near
the end of the Mesozoic eras.” H. 8, W.

6. Pre-Cambrian Igneous Rocks of the Fox River Valley,
Wisconsin ; by S. Werpman (Wis. Geol. and Nat. Hist. Surv.,
Bull. UI, Science, Series 2).—This is a geologic and petrographic
description of three areas of igneous rock, which appear to have
been islands during the deposition of Cambrian sediments. They

consist of an ancient rhyolite, a rhyolite changed to gneiss and a

; ;
Y
af

Tews. whe le en we

ay po ee

Y
i

=
=
-
;
2
my
z
i.
tt
>
<
re
i ,
‘
‘ ae
=

Geology and Natural History. 399

granite, which probably represent phases of a parent magma.

The various changes induced in these rocks through mechanical
deformation in two of the areas and through chemical alteration
alone in the third have been carefully studied and many features

of interest are developed. L. V.P.

7. West Virginia Geological Survey, vol. 1; by I. C. Wurrs,

State Geol. (Morgantown, West Va., 1899, 8°, 392 pp-)—In addi-

tion to the administrative report this volume contains a discussion
of magnetic declination for the state with description of meridian
monuments by R. U. Goopz. The State Geologist himself con-
tributes a list of levels above tide and a particularly valuable
report on the petroleum and natural gas of the state. This is
treated both from theoretic and practical standpoints and will be
found of great service to those engaged in a study of these
products. It is to be hoped that the state will continue the sur-
vey and permit the publication of other valuable material which
has been accumulated. . 1d fe at

8. A new and interesting olivine-melilite-leucite rock.—The
rock of one of two small volcanic cones discovered by E. Clerici
at San Venanzo in Umbria, about half way between Orvieto and
Perugia, has been described by V. Sapatini in a note (dated
September, 1898) in the Bolletino del Reale Comitato Geologico
for 1898, asa leucitic melilitite with olivine. Olivine occurs as -
phenocrysts, and the groundmass is composed of melilite, leucite,
black mica, and a little pyroxene, nephelite and magnetite, feld-
spars being absent. The melilites are zonally built, the interior
being optically positive and the exterior negative. Sabatini pro-
poses for this type the name Venanzite.

At the meeting of the Berlin Academy of Sciences held Feb. 9,
1899 (Sitzungsberichte, vil, p. 110, 1899), Prof. Rosenbusch
read a paper on “ Kuktolite, a new member of the theralitic effu-
Sive magmas.” ‘This rock is identical with that of Sabatini, com-
ing from the same locality, and his description is practically the
same though more detailed, and with chromite and perofskite
added to the list of component minerals. An analysis gave:

SiO, 41°43, TiO, 0:29, Al,O, 9:80, Fe,O, 3°28, FeO 5°15, MgO
13°40, CaO 16°62, Na,O 1°64, K,O 7°40, H,O 1:11 = 100°12.
P.O.none. Sp. gr. = 2°758.

He compares it with the madupite of Cross, leucitites, and
leucite and melilite basalts, all of which belong to his theralitic
magmas. The name is derived from evxrés, desired, since it
is “a desired example for the validity of my representation of
the intimate connection of the eruptive rocks throughout the

world.” Sabatini’sname Venanzite antedates Euktolite by about

five months and on the grounds of priority is unquestionably the
one by which this interesting type should be known, especially
since it is formed from the type locality, in accordance with
modern usage in petrography, aud does not involve the tacit
acceptance of any theory, however generally recognized.

H. 8S. W.

4.00 Scientific Intelligence.

9, Deuxieéme Mémoire sur les Algues Marines du Groenland ; .
by L. Koztperur Rosenviner. Meddelelser om Gronland, xx,
pp. 1-125, pl. i and 25 figs., 8°. Copenhagen, 1898. Om Alge-
negetationen ved Groniands Kyster; ditto, pp. 131-242.—The
paper first named is an important supplement to the author’s
Groenlands Havalger published in 1893. The material studied
included the collections made by N. Hartz, principally at Scoresby
Sound during the expedition to Kast Greenland under Lieut. ~
Ryder and those made by A. Jessen and C. Ostenfeld in West ~
Greenland, besides a few smaller collections. Of the alge of the
eastern coast, which had been less thoroughly explored than the
western, only 32 species had been previously enumerated, but in
the present work the number is estimated at 82 including several —
new species. The number of species previously known on the
west coast was 143, which number is now increased to 167 includ-
ing in the enumeration the species reported by Kuckuck from
Umanak collected by Vanhoffen. The Lithothamnia were
revised by Foslie. Among the novelties are three new genera,
Ceratocolax, Dermatocelis and Arthrochete, each represented by
a single species of parasitic habit. The paper is well illustrated
and the notes are full and suggestive. The second paper by Dr.
Rosenvinge is a valuable contribution to our knowledge of the
distribution of the marine floras of the North Atlantic, which
gives details as to the temperature and other factors which affect
the growth of marine plants and tables showing the relations of
the Greenland flora to those of northern Europe and the Atlantic
coast of North America. In spite of the remoteness and inclem-
ency of the Greenland coast, thanks to the publications of Rosen-
vinge, Kjellman and other Scandinavian algologists, its marine
flora is more accurately known than that of any other portion of
the Atlantic coast of America. W. G. F.

10. Monographie des Caulerpes ; by Mme. A. W2xBER-VAN
Bossg, Ann. Jardin Bot. Buitenzorg, xv, pp. 248-401, pl. xx—
xxiv.—In a short introduction the author gives a summary of
what is known regarding the cell structure with notes as to the
probable fructification of the species of this large and peculiar
genus, for, in spite of their abundance and wide distribution in
all warm seas, no one has as yet succeeded in detecting with cer-
tainty their method of reproduction. The monograph is essen-
tially a systematic study of the species for which the author is
especially well qualified, since her excursions to the tropics have
given her abundant opportunity for observing the Caulerpae in a
living condition, while, on the other hand, she has been able to
examine the dried specimens of practically all of the large algo-.
logical herbaria. The subdivisions of the genus adopted are
those of Prof. J. G. Agardh, but the number of species recog-
nized by Mme. Weber is considerably smaller than in other
works on the genus. In thus uniting the older described species
the author, we think, has shown good judgment, and her excellent
presentation of the subject gives a much better view of the genus

ae ee

Miscellaneous Intelligence. 401

as a whole than has been possible to obtain hitherto. Especially
in her treatment of the Thuyoideae and Claviferae is the author
to be commended. One is at first struck by the very large num-
ber of species reduced to synonyms under such species as C.
cupressoides and C. racemosa, which latter name replaces C.
clavifera as generally used hitherto. But one who has had the
opportunity of examining large sets of specimens can hardly fail
to approve the specific limitations adopted by the author, for not
only are the species themselves very variable but, also, one more
than- suspects that many of the species previously described
owed their assumed characters to some accident of preparation.
The arrangement of the present monograph seems to us to be the
natural one and one which is likely to be generally adopted.
W. G. F.

11. De dispositione Delessericarum ; by Prof. J. G. AGarpa,
pp. 239, 8°, Lund, 1898.—This volume, which forms the third
part of the third volume of the classic Species Genera et Ordines
Algarum, of which the first volume appeared in 1848, is a rare
illustration of continued scientific activity in a field in which the
writer has been a master for more than sixty years. To Amer-
ican algologists it is of special interest, as it includes a revision
of the species of Nitophyllum, Delesseria and related genera from
our Pacific coast which were originally described in Agardh’s
Epicrisis and subsequent memoirs, to which are added Nitophy/-
lum macroglossum, LV. stenoglossum, N. marginatum, N. Far.
lowianum, Neuroglossum lobuliferum. 'The new genus Erythro-
glossum with five species includes Delesseria Woodii and D.
Californica, and Apoglossum includes PD. dicipiens, all three
Californian species. The new genus Calloseris is founded on a
single species, C. Halliae, from Florida. W. G. F.

IIL. MiscELLANEOUS SCIENTIFIC INTELLIGENCE.

1. National Academy of Sciences.—The spring meeting of
the National Academy was held in Washington from April 18 to
20. The following gentlemen were elected members: Charles E.
Beecher of New Haven; George C. Comstock of Madison; Theo-
dore W. Richards of Cambridge; Edgar F. Smith of Philadel-
phia; Edmund B. Wilson of New York. The list of papers
accepted for reading is as follows:

W. K. Brooks and CASWELL GRAVE: Ophiura Brevispina.

A. Hau: The shadow of a planet.

A. AGAssiz: (n the Tanner deep-sea tow net. On the diamond and gold mines
of South Africa.

A. AGAssiz and A. G. MAYER: On the Acalephs of the East Coast of the
United States.

E. C. ANDREWs: On the limestones of Fiji.

W. McM. WoopwortuH: On the Bololo of Fiji and Samoa.

CuHAs. D. WatcoTtr: Progress in surveying and protection of the U.S. Forest
Reserves.

4.02 — Setentrifie Intelligence.

GIFFORD PrncHot: The work of the Division of Forestry, Department of Agri-
culture. ae

H. S. Pritcuretr: The resulting differences between the astronomie and geo-
detic latitudes and longitudes in the triangulation along the Thirty-ninth Parallel.

A. GRAHAM BELL: On the development by selection of supernumerary mam-
mee in sheep. Onkites with radial wings.

S. Newcoms: Remarks on the Work of the Nautical Almanac Office during the © $

years 1877--98 in the field of Theoretical Astronomy.

W. K. Brooxs and L. EH. Grirrin: Exhibition of specimens of Nautilus pom-
pilius. oe

The third memoir of volume vili of the Memoirs of the
Academy has recently been issued. The title of the paper is:
General perturbations of Minerva (93), by Jupiter, including
terms only of the first order with respect to the mass, together
with a correction of the elements, by W. 8. Hichelberger, Ph.D.

2. Annual Report of the Board of Regents of the Smithsonian
Institution, to July, 1897 ; pp. xlvii, 686. Washington, 1898.—
The Report of the U.S. Natural History Museum, under the
direction of the Smithsonian Institution, for 1896, was noticed in
the last number, and we have now issued the Smithsonian Report
for 1897. The early portion of this is devoted to an account of
administrative matters, by the Secretary, Professor 8. P. Langley.
This occupies some eighty pages. The remainder of the volume,
or General Appendix, contains reprints of a well-selected series
of papers on a very wide range of subjects; these include
memoirs on astronomy, physics (X-rays, ete.), chemistry and so
on through to archeology. The volume closes with an obituary
notice of George Brown Goode by S. P. Langley, and one of
Francis Amasa Walker by G. F. Hoar and C. D. Wright.

3. Harper's Scientific Series.—The recent additions to the
series of Harper’s scientific memoirs include Volume III, on
Roéntgen Rays, edited by Professor Grorce F. BarkER of
Philadelphia, and Volume IV, on the Modern Theory of Solution,
edited by Dr. Harry C. Jones of Johns Hopkins University.
The first of these contains the original paper by Réntgen, and
also others by Stokes and J. J. Thomson. Each of the papers is
followed by a biographical sketch of the author, and the bibli-
ography of the subject is added at the end. This subject is one
of very great interest at the present time, and indeed this state-
ment may also be made about Volume IV on the Theory of Solu-
tion. This latter contains papers by Pfeffer, Van *t Hoff,
' Arrhenius, and Raoult, with the usual useful editorial matter.

OBITUARY.

Dr. Gustav WirEpEMANN, the distinguished physicist, for
twenty-two years (1877-1899) editor of the Annalen der Physik
und Chemie, died on March 23 at the age of seventy-three.

45 specimens, ne received, are acknowledged to be
the best collection of crystallized specimens of this rare
metallic mineral ever placed on sale. ‘The crystals range
from 4 to 2 inch, averaging + to 2 inch; some of them
are intensely brilliant, others comparatively dull. Prices
for good cabinet sizes, 25c. to $3.00; a few museum-size
specimens at $2.00 to $6.00, and one very good group,
6x5x34 inches. of stout 4 to # inch crystals, $12.50.
The best cabinet specimens, $1.50 to $3.00.

NEW FINDS IN MONTANA AND IDAHO.

A Vesuvianite. Brilliant, yellow-brown erystals of rare development, in blue
' ealcite and light green pyroxene. Sizes 2x2 to 24x34. Prices, 15c. to Tic. 26
_ specimens. ; ,
» Wanadinite. A few attractive groups of brown to gray crystals, 25c. to
$1.00.
: Pyromorphite. Groups of light, yellowish green crystals, 25c. to $1.00.

CRYSTALLIZED COVELLITE.

ep From an out-of-the-way locality in Colorado, a few specimens have reached us
| ofa crudely crystallized Covellite. 25c. to $1.00.

RECENT FINDS IN UTAH. age ts.

“ Aurichalcite. Beautiful specimens at very low prices, lic. to 75c. Been 2k ee
_ blue-green, crystalline Smithsonite from same locality, 10c. to 35c. ited’
' __ Clinoclasite. Only six specimens, but three of them are uncommonly fine, E> Sea

$1.50 to $3.00; the others are good, 50c. to $1.00. see

ei: Anglesite. 22 groups of exceedingly brilliant crystals, the best seen in New een.
_ _ York for at least two years; priced for quick selling at only 2ic. to $1.50. meee Be
4 Scorodite, « few exceedingly choice specimens, 50c. to $1.25; Brochantite, ae
_ rare forms, 25c. to 75c.; Jarosite, excellent, 50c. to $1.00; crystallized Tyrol- . “gh ‘i
' ite, only three specimens, 50c. ani 75c. 29 good Olivenites ‘of various styles, eae
= 25c. to 75c. : 5
aS Sey ; oe | «etree
Be EXTRA STRONG LODESTONE. Fok PES
~ _110 pounds of the strongest Lodestone we have ever had. It is priced higher, 5 B50 :
po af course, but choice specimens cost only 15c. to 75c. and they are sure to please, : Rs C4
; FRANCKEITE AND CYLINDRITE. cate

100 tine large (!) specimens of these two new and rare Bolivian tin minerals. PE ioe Ms

No more are likely to come to this country except through our house. Prices for pe

_ pieces averaging 2x2 up to 24x4 inches, 0c. to $3.00. =

Kt ; ae :

a GORGEOUS POLISHED LABRADORITE! Be a

- We now have the finest assortment of specimens of polished Labradorite ever ‘Sae ‘l

seen in America. Cut on scientific principles, these slabs show far greater and eee" w"

_ more brilliant play of colors than is commonly seen. Rich yellow and coppery pee

___ colors, and strikingly handsome mottled specimens aboupd. Slabs up to 11x6 fore.” ie

__ inches in size, 50c. to $3.50. a : oe ai
eo oe i ars
_ MANY OTHER RECENT ADDITIONS ARE NOT MENTIONED. 12 ES eet
' _ 124-page Illustrated Catalogue, describing every mineral, 25c. in paper; 50c. in . SN
ee Pere, cloth. ; : ; oo
44-page Illustrated Price-Lists, also Bulletins and Circulars free. an
, Wane)

i GEO. L. ENGLISH & CO., Mineralogists,
oe REMOVED TO
‘aa 812 and 814 Greenwich Street (S. W. Corner of Jane Street), New York City.

Ss

CON TEM To

Art. XXXIV.—Some Experiments with Ene

Gases; by: W. G. Mixrer .-2-3oe eee et)
XXXV.—Hypothesis to explain the: partial non-expl siv
Combination of Explosive Gases and Gaseous Mix

Mexico; by H.-S, -WiLLIAMS.< be
XXXVII. —Origin of Paleotrochis; by J..8. Divee sae
XXX VIII.—Association of Argillaceous Rocks with Quartz
Veins in the Region of Diamantina, Brazil; by OAS oe
ah MRE eek ee 2 Dal. Seg ene
| XXXIX.—Goldschmidtite, a New Mineral ; by W. H. Hozss
XL.—Hydromica from New Jersey; by EF. W. CLARKE and
N.H. ‘DaRTon.. 1... 225 Li ee
XLI.—Powellite Crystals from Michigan; by C. Pavacue. |
XLIL—Volatilization of the Iron Chlorides in Analysis, |
the Separation of the Oxides of Iron and Aluminum ; ; by ee
F. A. Goocu and F. S. Havens... ae
XLIII.—Descriptions of imperfectly known and new Actini-
ans, with critical notes on other species, V3 Dy ae & ,
Vurrite. Bi a Le

y
XLV. lerady of Saat American Fossil One I:
by G. R. Wieianp. (With Plates VIILX)__ 2 m

SCIENTIFIC INTELLIGENCE.

CaJORI, 394. —_Blectrolytic papell A A.C. SWINTON: Meaunreres 0:
electrical oscillations, W. K6x1c, 395.—Electric waves on wires, W. D
m@E: Absorption of the X-rays by air: Optischen Tastee der

Fuess, C. Less, 396. sie

Geology and Natural Histor y—Note on a Bridger Kocene Carnivore, 0. c. Mt
Age of the Franklin White Limestone of Sussex County, N. 7 J. Bo We
and A. H. Brooks, 397.—West Virginia Geological and Hconomic Surye
Alabama Geological Survey, W. B. Put~irps: Development of Lytoceras
Phylloceras, J. P. Smitu: Pre-Cambrian Igneous Rocks of the Fox River —
ley, Wisc., S. WEIDMAN, 398 —West Virginia Geological Survey, I. C. Wu
New and interesting olivine- -melilite-leucite rock, V. SaBaTrNI, 399.—Deux uxi
Mémoire sur les Aleues Marines du Groenland, L. K. ROSENVINGE:
graphie des Caulerpes, A. W. Bossn, 400 _—Dispositione Dele
AGARDH, 401. 8

Miscellaneous Scientific Intelligence—National Academy of galdenes 401
Report of the Board of Regents of the Smithsonian insta to phere ;
Harper’s Scientific Series, 402. B

Obituary—Dr. GusTAV WIEDEMANN, 402.

‘ THE

Epitor: EDWARD &8. DANA.

: ASSOCIATE EDITORS

IL P. BOWDITC JH anp W.-G. FARLOW, or CAMBRIDGE,

| | Prosssons A. E. VERRILL, HENRY 8. WILLIAMS, an»
Vs De Be V. PIRSSON,, oF New Haven,

Prorussorn GEORGE F. BARKER, or Putapeten,
Prorrssor H. A. ROWLAND, oF Batrimors,
Mr. J. S. DILLER, or Wasnineton.

FOURTH SERIES.
VOL. VII-[WHOLE NUMBER, CLVII.]
No. 42.—JUNE, 1899.

WITH PLATE XI.

NEW HAVEN, CONNECTICUT.
1899. tae

TUTTLE, MOREHOUSE & TAYLOR, PRINTERS, 125 TEMPLE STREET.

Published monthly. Six dollars per year (postage prepaid). $6.40 to Roters
ign subscribers of countries in the Postal Union. Remittances should iad
me. either by money orders, registered letters, or bank checks. vou

QUEEN STA

_ PRECIOUS OPA

ae loa Prices. |

A large collection of this beautiful and popular stone, purchased by 0)
tralian collector, was opened for sale May Ist. A description of so well no
gem seems almost unnecessary, but the purchase of a specimen will ex
use of superlatives in referring to it. Of few things is it as easy to say fine:
as of the happily named Noble Opal. An exquisite play of delicate colors,
perhaps bolder and grander flashes of varied lights, spread showily over a broz
surface of rock, have won a reputation for this gem among all others
prominent colors are green and blue, with splashes of red and purple. Oft

~ opal has considerable “depth, yielding gem material of the first quality. eh
no “washed out” look about such specimens. They present a natural dis
combined and brilliant color, in varied Cuanees that are alike the pe 0
and counterfeiter.

in a brown ee matrix which is often of a flinty texture, serving as a
but excellent setting. Our present prices are most reasonable, barely CG
the actual gem value. The best generally range from I to 2 inches d
and are priced at $2.00 to $8,00. <A few at $10. 00 to S15. 00, including .
specimens of unusual beauty. Gem fragments and pieces ‘of matrix
charming bits of color, as low as 15e, to $1. 50. They are infinitely snperiol
Mexican, and should be in every collection. A good stock of cut onEiste

$8.00 per carat.
CROCOITE.

NEW CONSIGNMENT. 50 PER CENT. REDUCT ON.

has met with phenomenal success. The first stock of good specimens ae
exhausted, but is replenished by a small series just received. Although
find they were but recently bought in by our collector, and at such ter
enable us to cut former rates in half. Gorgeous orange- “red masses of crystals
ferromanganese with lustrous and perfectly terminated crystals of varied ha
at $1.00 to $4.00. Several at $5.00 to $8.00. Detached crystals and §
_Inatrix specimens, 15c. to 75c. Pure crystal fragments, 50c. per oz.

Mesolite, Phacolite, Phillipsite, Ferrocalcite, Aragonite and «
minerals from the Collingwood Quarries, near Melbourne, reached us in the ;
consignment. See Catalogue. oe Spe:

Many other New Arrivals are mentioned in our 16 pp. Illustrated ‘Su
ment, including Crystal and. Pound Material addenda lists, New Species, -
wine Collection, ete., etc., mailed free. se

Illustrated Collection Catalogue, 64 Pp. Prospectors and educati
setS . Free. .

Illustrated Catalogue of Cabinet Specimens, 126 pp. “Free.

Is" MINERALS PURCHASED. —

WARREN M. FOOTE, Manager. hag

Minerals for Scientific and Educational Purges ae GE
1317 Arch Street, Philadelphia, Bo U. J.

ESTABLISHED 1876:

mM,

/
:

TH:

AMERICAN JOURNAL OF SCIENCE

[FOURTH SERIES. ]

OTHNIEL CHARLES MARSH.

Amone the leading men of science in America, Professor
Marsh was unguestionably one of the best known, and had one
of the strongest personalities. The world-wide reputation he
enjoyed, however, is not altogether attributable to the particu-
lar department of research in which he stood without a peer,
for, added to his attainments in Vertebrate Paleontology, he
possessed an unusual number of mental qualifications in other
lines, as well as marked personal characteristics which made
him known and felt where his science could never reach. His
fame will undoubtedly rest on his work among the Fossil
Vertebrates. Nevertheless, his energy and attainments in other
directions were sufficient to have made for him a permanent
record.

The nearness of the perspective at the present time renders
it difficult properly to individualize and accord the true rank
to the many important discoveries Marsh has made. He
brought forth in such rapid succession so many astonishing
things that the unexpected became the rule. The science of
Vertebrate Paleontology could not assimilate new material so
fast, and it will be years before the true significance and bear-
ing of much that he has done will be understood. The con-
stant:stream of vertebrate riches which, from 1868 to 1899,
flowed into the Yale University Museum from the Rocky
Mountain region had a similar bewildering effect on Marsh, for

Am. Jour. So1.—FourtH Series, Vou. VII, No. 42.—JunzE, 1899.

404. Othniel Charles Marsh.

it was impossible for him to do more than seize on what
appealed to him as the most salient. The work of the hour
was to him of prime importance, whether it was for the deter-
mination of a new order of mammals or a new cusp on a
tooth. Still, he seems to have had a just conception of relative
values, for it will be found that he plucked the most luscious
plums from the paleontological tree, and left chiefly the smaller
or unripe and imperfect fruit untouched.

Another element in his success was seen in the improve-
ment he made in the methods of collecting, preserving, and
developing vertebrate fossils, so that even forms long known
only from fragmentary remains were represented in his collec-
tions by almost complete specimens, presenting nearly the
same degree of novelty shown in forms actually new.

In illustration of this, the Brontotherids, Ceratopsia, and
the Mosasauria furnish excellent examples. Prout, in 1846,
described, as Paleothervum, the fragment of a lower jaw from
the Miocene of Nebraska, but Marsh first showed the affinities
and range of forms in the group, through his splendid restora-
tion of Brontops and the description of a number of allied
types from nearly perfect material. Cope, in 1875, figured
some pieces of bone of unknown relationships, which long
remained in the paleontological scrap-basket.* Marsh, by his
descriptions of the marvelous series of genera and species
belonging to the Ceratopsia, demonstrated what these reptiles
really were, and gave to science a nearly complete knowledge
of one of the most bizarre monsters known. ‘The first
Mosasaur was obtained in Holland previous to 1785. It
remained imperfectly known for nearly a century, when Marsh,
by his contributions to its anatomy, made possible a clear
understanding of its structure and affinities. Im the same way
it could be shown that to many old descriptions of genera and
species based upon single teeth, he was enabled to add a
knowledge of the remainder of the animal. Not only did he
thus contribute the missing information in regard to many
previously described forms, but he brought out a host of
entirely new types, and made his science one of the most com-
plete exponents of the doctrine of evolution.

* Polyonax.

Othniel Charles Marsh. 405

As a collector, Marsh was seen at his best, and the collec-
tions he amassed during his forty-five years and more of
activity in this direction form a lasting monument to his per-
severance and foresight. A person with means and inclina-
tion may be supposed to have the necessary qualities for
accomplishing his aims, whether they are first editions, auto-
graphs, or fossils, but had Marsh possessed no further qualifi-
cations than these, the results of his collecting would fall far
short of what he really attained. He not only had the means
and the inclination, but entered every field of acquisition
with the dominating ambition to obtain everything there was
in it, and leave not a single scrap behind. Every avenue of
approach was made use of, and cost was often a secondary con-
sideration. The nine-tenths, when attained, were only an
additional stimulus for securing the remaining one-tenth. Of
course, this ideal of completeness was often impossible of
accomplishment, and yet it served to bring to the Yale Uni
versity Museum collections which are unique from their rich-
ness and extent.

In making an estimate of his character, it must not be
forgotten that he developed wholly without the influence of
family and home ties, which in most men profoundly mark
their mature life. Self-reliance is probably the strongest trait
fostered by the absence of immediate family connections.
This, Marsh possessed to an extraordinary degree, and it natu-
rally led to a self-centering of his life and ambitions. Out of
it came, also, an absence of the complete exchange of confi-
dence which normally exists between intimate friends. Even
where perfect confidence existed, he seldom revealed more
about any particular matter than seemed to him necessary or
than the circumstances really demanded. As a friend, he was
kind, loyal, and generous. As a patron of science, he has
seldom been equaled. Honest work in any department
appealed to him strongly, and he was ever ready with aid and
counsel, even at the expense of a personal sacrifice. His dis-
position was a most happy one, and he was always keenly
appreciative of the humorous and ludicrous and fond of relat-
ing amusing experiences and anecdotes. The sunny side of his

406 Othniel Charles Marsh.

character was nearly always uppermost, and the consideration
of subjects of the greatest gravity was enlivened by constant
sparkles of wit from his exhaustless store.

He was normally restive under restraint, and met all opposi-
tion with power and fearlessness. Having practically created
the modern science of Vertebrate Paleontology in America, he
resented any encroachment upon the particular fields of research
in which he was engaged. This attitude frequently devel-
oped feelings of hostility in other investigators, and often
alienated him from co-workers in his department of science.
Nevertheless, he labored faithfully for the truth as revealed in
his work, and was ready to change opinions and eae
statements whenever facts seemed to warrant it.

His esthetic sense was highly developed, and could be seen
in the artistic care he bestowed upon his publications, but more
especially on his home. His grounds are a model of landscape
gardening. He delighted in his collections of modern paint-
ings, the cultivation of orchids, and above all in the subtleties
of Japanese art.

The world was not slow to recognize his contributions to
knowledge, for during his lifetime he received a large number
of tangible evidences of distinguished consideration in the way
of academic and scientific honors, medals, and membership in
learned societies.

In 1886, he received the degree of Doctor of Laws from
Harvard University, and in the same year the honorary degree
of Doctor of Philosophy from the University of Heidelberg.
He occupied the chair of Paleontology in Yale University from

ee a

1866 to the time of his death. He was Vertebrate Paleontolo- 4

gist to the United States Geological Survey, and Honorary
Curator of Vertebrate Paleontology in the United States
National Museum.

He was President of the American Association for the
Advancement of Science in 1878, and of the National
Academy of Sciences from 1883 to 1895. As a presiding
officer in the National Academy, he exercised the same amount
of care that he bestowed upon his private affairs, and was an
active and efficient leader.

ibis 9s,

Othniel Charles Marsh. 407

In 1877, he was the recipient of the first Bigsby Medal
awarded by the Geological Society of London, in recognition
of his important labors on the Vertebrate Paleontology of the
western territories of the United States. In 1898, the highly
valued Cuvier Prize was given him by the French Academy,
as one of the most able continuators of the science of which
Cuvier had laid the foundations.

Prominent among the various societies of which he was a
member may be mentioned :

The National Academy of Sciences: Institute of France;
Royal Academy of Sciences, Brussels; Royal Bavarian
Academy of Sciences, Munich; Royal Academy of Sciences,
Bologna; Royal Danish Academy of Sciences, Copenhagen ;
Royal Irish Academy; Geological Society of London; Geo-
logical Society of Germany ; American Philosophical Society ;
Academy of Natural Sciences, Philadelphia; Zoological
Society of London; Société Impériale des Naturalistes, Mos-
cow ; Geological Society of America, etc., ete.

Few men have contributed more to THE American JOURNAL
or Screncre than Professor Marsh. Nearly all his discoveries
in science were first announced here, and it is the storehouse
of most of his best work.

The subject of the present sketch was born near Lockport,
New York, October 29, 1831. His parents were Caleb and
Mary Peabody Marsh, formerly of Danvers (now Peabody),
Massachusetts. His early education was obtained in the
schools of Lockport and at the Wilson Collegiate Institute,
Wiison, New York. A residence in a region rich in minerals
and fossils is apt to attract the attention of a youth possessing
healthy intelligence, and young Marsh soon shared his vacation
time between the normal pursuits of shooting and fishing and
the more unusual vocation of collecting minerals and fossils.
By the time he was nineteen years old, he had thus acquired
the taste for scientific subjects which was destined to grow and
dominate the remainder of his life.

In 1851, he entered Phillips Academy at Andover, Massa-
chusetts, and continued his studies there until graduation in

408 Othniel Charles Marsh.

1856. He immediately entered the freshman class in Yale
College, pursuing the regular classical course, and receiving
the degree of B.A. in 1860. Graduate courses in the natural
sciences were continued in the Sheffield Scientific School during
the two years following (1861-62). The long summer vacations
from 1851 to 1862 were occupied in collecting minerals and
fossils from New York, New England, and Nova Scotia. To
the latter region he made five trips during this interval, and
obtained much valuable experience and scientific material.
On his second visit (1855) he found some fossil vertebree in
the Coal Measures at South Joggins, representing a new and
important vertebrate animal (Hosaurus). This discovery
finally directed his studies into the channel which became his
life-work. At this time, however, his interests were about
equally divided between invertebrate paleontology and miner-

alogy, and it is worthy of note that his first scientific paper:

was published in tats JourNAL in 1861, under the title “ The
Gold of Nova Scotia.”

The description of Hosawrus did not appear until 1862,
seven years after its discovery. Even then it cannot be said
that he had developed a strung liking for vertebrate paleon-
tology. This closes the account of his student life in American
schools. ae

The next three years were passed in study abroad, in the

universities of Berlin, Heidelberg, and Breslau. He attended —
lectures and took special courses with H. Rose, G. Rose,

Ehrenberg, Peters, Roemer, Grube,’ and Geeppert. The vaca-
tions were occupied, as before, by geological excursions. He
visited the most important localities in Europe, and obtained
extensive collections. His official connection with Yale Col-
lege began by his appointment, in 1866, to the chair of
Professor of Paleontology. This title he held in high esteem,
as it was the first established either in this country or else-
where. |
After attending the meeting of the American Association
for the Advancement of Science at Chicago, in 1868, Marsh
went as far west as Nebraska and Wyoming, along the route
of the Union Pacific railroad, then just opened. This trip
gave him a foretaste of the inexhaustible fossil riches of the

a
by
Z
”
a
j
:
-
.
SS
7
‘=
‘
-
q
¥
*
a
.
>
;
4
a

Othniel Charles Marsh. 409

Rocky Mountain regions, and thenceforth his energies were
mainly devoted to their exploration. Scientific expeditions to
the western country were undertakings of considerable magni-
tude in those early days. There was but one railroad in the
United States across a region measuring fifteen hundred miles
square. White settlements were sparse and remote. Most of
the country was unmapped, and with the exception of a few
transcontinental trails, almost the whole western half of the
continent, save the regions bordering the Pacific, was a
boundless expanse of unknown arid plains, mountains, and
valleys. Added to these conditions were the indigenous tribes
of war-loving Indians, hostile to the whites. Under such cir-
cumstanees, travel was slow, difficult, and dangerous. It was
necessary to have an escort of soldiers and guides, experienced
in western life and Indian warfare.

The first Yale Scientific Expedition was organized and engi-
neered by Marsh in 1870. The party consisted of thirteen
persons besides the officers and men of the military detach-
ments who escorted them from various military posts along
the route.* They explored the Pliocene deposits of Nebraska
and the Miocene of northern Colorado, then crossing into
Wyoming they made collections in the Eocene (Bridger Basin),
and passing south discovered a new Eocene basin in Utah
(Uinta Basin). At each of these places many important finds
were made. The party next visited -California, where minor
collections were obtained from the Pliocene. Returning, they

* Members of the Yale party were O. C. Marsh, C. T. Ballard, C. W. Betts.
A. H. Ewing, G. B. Grinnell) J. W. Griswold, J. R. Nicholson, C. McC. Reeve,
J. M. Russell, H. B. Sargent, J. W. Wadsworth, E. Whitney, Jr., and H. D.
Ziegler. The escorts consisted of :—

From Fort McPherson, Nebraska.—Commanding officer, Gen. Eugene A. Carr.
Lieuts. Bernard Reilly, Jr., and Earl D. Thomas, in command of escort, 5th Cav-
alry; Buffalo Bill and Major Frank North, guides; and two Pawnee Indian
scouts (“‘Lahurasoc” and ‘‘Tuckatelous”’).

From Fort D. A. Russell, Wyoming.—Commanding officer, Gen. John H. King.
Capt. Robert H. Montgomery and Lieut James McB. Stembel, in command of
escort, 5th Cavairy.

From Fort Bridger, Wyoming.—Commanding officer, Major R. S. LaMotte.
Lieut. W. N. Wann, in command of escort, 13th Infantry: Mexican guide (‘‘Joe
Talemans ”’).

From Fort Wallace, Kansas —Commanding officer, Gen. Henry C. Bankhead.
Ed. Lane, guide; Lieut. Charles Braden, in command of rescue troop.

410 Othniel Charles Marsh.

spent some time exploring the Cretaceous beds of western
Kansas, so rich in the remains of aquatic reptiles, and now
famous for having furnished the first toothed birds and Ameri-
can toothless flying reptiles.

The second, third, and fourth Yale Scientific Expeditions
(1871, 1872, 1873) were modeled after the first. New regions

in the West were visited, and extensive series of remains of —

extinct animals were obtained. Coincident with these discov-
eries, Marsh published frequent scientific papers describing and
illustrating the more important forms, and paleontological
literature was enriched by the addition of more startling and
wonderful types of animal life than had been hitherto known
from the rest of the world.

Owing to Indian outbreaks and a general uneasiness in the
West, no regular expedition was organized in 1875. Late in
the fall, however, Marsh went to the Bad Lands of Nebraska
and Dakota accompanied by an escort from Fort Laramie to
the Red Cloud Agency. The consent of the Indians was
deemed necessary to search for fossil bones in their country.
A treaty was obtained with difficulty and then assistance was
withheld. Nevertheless, with great hardship owing to extreme
cold, the party succeeded in reaching the desired region, and
made important discoveries, among which numerous remains
of the gigantic Brontotheride are the most noteworthy.

It was at this time that he became aware of the frauds prac-
ticed upon the Indians by the agents of the Government, and
the way the Government was in turn defrauded through their
misrepresentations. He promised Red Cloud to bring the
matter before the President for redress. This was done with
signal success, resulting in the complete routing of the Indian

Ring, and ‘he eal of the Secretary of the Interior as

well as in his political death.

The rapid settlement and development of ihe West rendered
it no longer necessary to fit out expensive expeditions, espe-
clally as many of the localities were easily accessible by rail-
road. Therefore, after 1876, local collectors and small parties
were employed in continuing the work of collecting fossils so
successfully begun by the Yale Scientific Expeditions. Nearly
every season, however, Marsh visited the localities where work

ne
ai
-
a
ie:

ee)

5 iid tal ee hin ee il aes is iia ae fault

i.

-
_—s

ee +.

%
3
:
‘

Othniel Charles Marsh. 441

was being carried on, and some time each year was spent in
reconnaissance for new fields of labor.

The right wing of the Peabody Museum was completed in
1875, the means having been furnished by Mr. George
Peabody largely through the influence of his nephew, Profes-
sor Marsh. It was to his uncle, also, that Marsh was indebted
for his educational advantages and for his private fortune.
The old Yale Cabinet had long been outgrown. The rooms
became so crowded that for years there was only space for a
chalk line dividing the different departments. The collections
which had been accumulating during so many previous years
found a commodious home in the new museum, and work
was resumed with great activity under more favorable condi-
tions than heretofore. Huxley’s visit in the following year
was a further stimulus to higher work, as is clearly evinced
in the celebrated Nashville address mentioned elsewhere.

The National Government had not altogether neglected its
opportunities for scientific research in the West during this
period, though the results in the way of substantial collections
were far inferior to those Marsh had obtained. For some time
previous to 1878, there were four separate surveys, two under
the Engineer Department of the Army and two others, exten-
sions of private expeditions, under the Department of the
Interior. In the reorganization ordered by Congress in 1878,
Marsh, as acting President of the National Academy of Sci-
ences, was the chief instrument in effecting a consolidation and
in defining the relations of the present United States Geologi-
cal Survey with the general Government and with the United
States National Museum. The wisdom of this change was at
once apparent, and the Survey is now often considered one of
the most economical, best managed, and productive depart-
ments of the Government.

After repeated solicitation and with promises of material aid
in the way of publication and collections, Marsh, in 1882,
accepted the appointment of Vertebrate Paleontologist to the
United States Geological Survey. This position he held to the
time of his death, although the field work for the survey was
terminated in 1892. His connection with the Survey gave him
increased facilities for publication and for prosecuting explora-

419 Othniel Charles Marsh.

- tions in the West. He successively projected the publication
of a number of large monographs on various groups of verte-
brate fossils. It is a great misfortune that but two of these
were ever finished by the author. The monograph of the

Odontornithes appeared in 1880, and that of the Dinocerata in

1885. The others were left in various stages of incomplete-.
ness at the time of his death. The proposed volumes treated
of the Sauropoda, the Brontotheride, the Stegosauria, Thero-
poda, Ornithopoda, Mesozoic Mammals, and the Ceratopsia.
Most of the investigations had been completed, a large part of
the plates and figures engraved, and preliminary descriptions
published, but the philosophical and phylogenetic problems are
largely untouched. The loss to science is greatest in the vol-
umes relating to Reptiles, especially the Dinosauria, for in
this subject Marsh stood as the sole possessor of an acute
and comprehensive knowledge of one of the most wonderful
ana difficult groups of vertebrates known. He planned his
life-work on the basis that immortality is here and not in the
hereafter. It seemed difficult for him to realize the limita-
tions of human existence and worldly accomplishment.

In the closing years of his life he had two ruling ambitions,—
first, to see the main building of the Museum erected, and,
second, the completion of his monographs. The accomplish-
ment of the first is imperative and would permit of the proper
care and display of the priceless treasures he has accumulated.
The attainment of the second would cancel his obligations to
science. Neither was realized. |

As one of the trustees of the Peabody Museum and as
Curator of the Geological Collections, Marsh performed his

chief duties in connection with Yale University. The final

transfer to the University, of all the collections he had aceumu-
lated, was made January Ist, 1898, and soon after the gift was
accepted by the Corporation. These collections are so exten-
sive as to merit particular attention, especially since they rep-
resent the most valuable part of the work of a lifetime, and
form the chief monument of one of Yale’s most noted men.
As expressed in the deed of gift, the collections comprise :

A ow = ates? = or A

Se ee

|
|
|
i
|

i. a ©

=

— . -

Othniel Charles Marsh. 413

1. The Collection of Vertebrate Fossils. This is the most
important and valuable of all. It is very extensive and contains
a large number of type specimens, many of them unique,
and is widely known from the descriptions already published.
In extinct Mammals, Birds, and Reptiles, of North America,
this series stands preéminent. The collection was pronounced
by Huxley, who examined it with care in 1876, to be surpassed
by no other in the world; and Darwin, in 1878, expressed a
strong desire to visit America for the sole purpose of seeing
it. Since then it has been more than doubled im size and
value, and still holds first rank. The bulk of this collection
was secured in western explorations, which were extended over
a period of nearly thirty years.

2. The Collection of Fossil Footprints. ‘These specimens
are mainly from the Connecticut Valley, and thus have a
special local interest.

3. The Collection of Invertebrate Fossils. This includes a
large amount of interesting material from many formations
and localities, both in this country and in Europe. Among the
series of specimens especially valuable may be mentioned
several thousand from the famous Mazon Creek locality in Ih-
nois ; a very extensive collection of Crinoids from Crawfords-
ville, in Indiana; the largest collection of nearly entire Trilo-
bites yet discovered ; and one of the rarest series of Silurian
Sponges known, including important type specimens.

4. The Collection of Recent Osteology. This is believed to
be one of the most complete collections in this country for pur-
poses of study. Special efforts have been made for many years
to secure the skeletons of rare existing vertebrates from every
part of the world, particularly of Mammals, Birds, and Rep-
tiles. The collection is especially rich in Anthropoid Apes.

5. The Collection of American Archeology and Ethnology.
This collection is replete in Central American antiquities, com-
prising several thousand, many of them unique. Among others
is the famous deZeltner collection from the same region, con-
taining a number of gold ornaments. The specimens from
Mexico are also of great interest, and the series is a repre-
sentative one. It includes the well-known Skilton collection.

6. The Collection of Minerals. This is a limited collection,
but contains many valuable specimens, among them probably

414 Othniel Charles Marsh.

the most interesting series known of Nova-Scotian Zeolites.
These were mainly collected by Marsh, before he was gradu-
ated at Yale, during six expeditions to Nova Scotia.

Besides the six main collections named, there are several
others of less value, which include fossil plants, casts of fossils,
geological specimens, and recent zoological material.

To these should be added the results of his last work in
endeavoring to increase the scope of the material in the Pea-
body Museum. For many years it was his desire to secure a
collection of fossil Cycads, and when the opportunity offered,
he embraced it with characteristic vigor, so that within the
last year and a half the Museum has received an amount of
material which in importance and quantity 1s second to none.

From their extensive and varied nature, these collections thus
presented to the University will long afford abundant material
for original investigations, and will ever attract to New Haven
specialists in Paleontology and Archeology.

Professor Marsh’s life was remarkably free from the petty
annoyances of poor health which so often interfere with human
comfort and ambitions. In the midst of his scientific work
and while making plans for the growth of the Museum, he
was suddenly overtaken by the malady which resulted in his
death. He died of pneumonia, on March 18th, 1899, in his
sixty-eighth year, after an illness of about a week. His work
as an investigator in natural science, his wonderful scientific
collections, and his munificence to Yale, are his legacies to
the higher education of mankind.

Although Marsh was an ardent collector in Archeology, he
published very little on this subject, and his paper (1866) on
an Ancient Sepulchral Mound near Newark, Ohio, is practi-
cally the only one. His three mineralogical papers, published
between 1861 and 1867, show the results of considerable labor
and careful investigation. They treat of the Gold of Nova
Scotia, a Zeolite mineral from the same region, and a catalogue
of the mineral localities of the maritime provinces of Canada.

In the field of Invertebrate Paleontology, he likewise was an
indefatigable accumulator of material, though after 1869 he

Othniel Charles Marsh. 415

published nothing in this department. Two papers presented
some Annelids considered as new, from the Jurassic of Ger-
many. Another showed the origin of the double lobe-lines in
Ceratites. His papers on American invertebrates comprised a
description of a new genus of Fossil Sponge ( Brachiospongia ),
a new form of Crustacean Trail from the Potsdam Sandstone,
and a note on color markings in Fndoceras. He also showed
that Paleotrochis and Lignilites were not of organic een
though the contrary had been previously supposed.

In the domain of Geology, his chief imterests lay in the
formations from which he secured important series of fossil
vertebrates. Probably his greatest geological discovery was
the Uinta Basin, an Eocene deposit of the eastern Uinta
Mountains. It was first visited in 1870. Having studied most
of the Tertiary lake basins in the Rocky Mountain region, he
gave, in 1875, a synopsis of their geological features. As a
natural result of studying Geology in Germany, he was much
impressed with the methods of marking the separate horizons
by means of some characteristic fossil. He believed the verte-
brates were the most sensitive time-markers, and therefore
endeavored to determine and limit geological horizons wholly
by fossil vertebrate remains. The inherent fault of this sys-
tem is that the vertebrates are not always the most highly
differentiated and specialized types in any given fauna, and it is
these qualities alone that can be safely employed in organic
chronometry. This method is usually of great value in fresh-
water deposits rich in vertebrate remains, but it can be seldom
used to advantage in marine sediments or in formations
containing a scanty vertebrate fauna. Thus, while the name
Equus Beds is very appropriate for a horizon in the Pliocene,
on account of the abundance of remains of fossil horses, the
same cannot be said of the term Eosaurus Beds as an equivalent
of the entire series of the Coal Measures, especially as but two
vertebree of this animal have ever been discovered. Geolog-
ical facts will be found scattered through many of his publica-
tions dealing principally with fossil vertebrates. One of the
latest problems to interest him was the age of the series of
variegated clays extending from Martha’s Vineyard south along
the Atlantic coast into Maryland. His investigations led him

416 Othniel Charles Marsh.

to refer them to the Jurassic, a formation which had been con-
sidered as absent in eastern North America.

There yet remains for consideration the real work of his
life,—his publications on the Fossil Vertebrates, and it is at
once evident, from a glance at the bibliography, that his chief
researches were upon the Reptiles, Birds, and Mammals.
There are three papers on Fossil Fishes, containing notices of
several new forms, but no real research in this class was ever
undertaken by him. The Amphibians also claimed but little
attention, and his observations on the metamorphosis of the
recent Siredon into Amblystoma, and two brief notices of
amphibian footprints in the penne and Carboniferous, com-
prise the whole.

It is with extreme hesitation and a sense of inadequacy that
the writer ventures to review, even in the briefest and most
superficial. manner, the pork which undoubtedly constitutes
the literary essence of his life-work. [Future investigators
alone can critically estimate the great mass of facts which
Marsh brought out and which he wove into the departments
of fossil Reptiles, Birds, and Mammals.

His most binpichiensire work, and in many ways the most
masterly, is the address delivered before the American Asso-
ciation for the Advancement of Science, at Nashville, in 1877.
In this paper, entitled the “Introduction and Succession of
Vertebrate Life in America,” he traced the introduction of the
various types of vertebrate life then known in America, begin-
ning with the lowest fishes and ending with man. The amount
of knowledge on the lower classes of vertebrates, including the
reptiles, was then too meager to,enable him to give more than
occasional hints as to their phylogeny. But his handling of
the Mammalia showed the clearest insight into the develop-
ment and affinities of many of the important types, and marked
him as a true philosopher.

A glance at the modern text-books of Geology and Paleon-
tology reveals how much America has done for the fossil ver-
tebrates in the three classes of Reptiles, Birds, and Mammals.
It will also show that Marsh contributed more than any other
investigator toward the prominence now accorded to the
American forms.

Othniel Charles Marsh. ALT]

His work on the Reptilia is not equally divided among the
various orders, for the Dinosauria claimed his attention above
all others. To this group he lent his best efforts, and he com-
passed it so thoroughly as to be its sole master. It seems only
necessary in this place to notice the complete restorations he
made of some of these remarkable animals. In this list are
included Anchisaurus, Brontosaurus, Laosaurus, Cerato-
saurus, Camptosaurus, Stegosaurus, Triceratops, and Clao-
saurus. It must be remembered that nearly all these animals
were of gigantic stature, some of them the largest land
animals yet known, and also that each restoration represents a.
number of separate investigations on the structure of the skull,
the limbs, the vertebrze, the pelvis, etc. In most cases, only by
this means was it possible to bring together gradually, part by
part, until the sum of the knowledge warranted a complete
representation of the skeleton. The material of many of the
genera he described is still in these various stages of progress,
awaiting new additions of portions yet unknown in order to
form a finished conception of the entire animal. His exten-
sive report on the Dinosaurs of North America, published in
1896, gave a synopsis of what he had accomplished up to that .
time, but as remarked elsewhere their philosophical treatment
he had reserved for his final monographs.

Probably, among the Reptilia, next in importance to his
work on the Dinosauria is that on the Mosasaurs. In this he
first announced the discovery of the dermal armor, the position
of the quadrate, the finding of the stapes, the columella, the
hyoid, the sclerotic plates, the quadrato-parietal arch, the malar
arch, the transverse bone, the pterygoids, the pterotic bone,
the sternum, the anterior limbs, the posterior limbs, the length
of the neck, and details of the pelvic region. Thus he con-
tributed a knowledge of some of the most essential characters
of the skeleton in this group. In other groups of aquatic rep-
tiles, he also brought out new genera and types of structure.
Prominent among these may be mentioned Laptanodon, a
toothless Ichthyosaurian. Marsh was the first to describe the
remains of fossil serpents in the western Tertiary deposits, and
likewise the first to discover the remains of: flying reptiles in
America. The latter were of unusually large size and remark-
able for the absence of teeth.

418 Othniel Charles Marsh.

The acquisition of a unique specimen of Pterodactyl from
the lithographic slates of Bavaria enabled him to supply the
long sought information regarding the wing and caudal mem-
branes. Notices of a number of new species of fossil Croco-
diles, Lizards, and Turtles, complete this survey of his work on
the Reptilia. .

Practically, most of the present knowledge of extinet bird-
life in America is contained in Marsh’s publications, which
include descriptions of numerous species, ranging from the
Jurassic to the Post-Phocene. Unquestionably, the one dis-
covery which is always foremost in men’s minds in a considera-
tion of his work is the-determination of an extinct order of
birds possessed with teeth. The study of the Dinosaurs and
Toothed Birds showed that one by one characters considered
as avian were likewise present in reptiles, and that many rep-
tilian characters were present in these primitive birds ; so that
at the end there did not seem much else besides feathers to
distinguish them. Marsh’s investigation of fossil birds led to
the publication, in 1880, of his first monograph, “ Odontor-
nithes: a Monograph on the Extinct Toothed Birds of North
America.” In this volume, he carefully figured and described
all the known types, and presented complete restorations of the
two leading genera, Hesperornis and Ichthyornis. He con-
cluded that birds most nearly resemble some of the small
Dinosaurs from the American Jurassic, and that both classes
originated at least as far back as the Trias or late Paleozoic,
in some sauropsid type.

A discovery which rivaled that of the Toothed Birds,
although not so wholly his, was the genealogy of the Horse.
Huxley and Kovalevski traced the equine branch through the
Pliocene to the Upper Miocene in Europe, but the true and
remote ancestry remained unsolved until the American types
were described by Marsh. He showed that a primitive and
diminutive polydactyl horse existed in the Lower Eocene, and
that from this type, by gradual and progressive change through
successive horizons of the Eocene, Miocene and Pliocene,
there had been evolved all the intermediate stages leading to
the modern horse.
| Next in importance and interest should be noticed the series
1) of papers culminating in the monograph of the Jinocerata,

a

:
E
if
\
;
;

rer we = §

Othniel Charles Marsh. 419

issued in 1886 by the United States Geological Survey. His
work in other groups of mammals is scattered through a large
number of separate papers, and contributions were made to
every known order. The Tillodontia comprise one of the
most remarkable of the types. Among others are the first
remains of fossil Primates, Cheiroptera, and Marsupialia,
known from North America. The Brontotheride and
Coryphodontia received considerable attention. A monograph
had been begun on the former, and restorations of a typical
genus of each were published.

One general conclusion of much significance was the out-
come of his researches on the Mammals. It was that the
Tertiary genera possessed very small brains. As a single
example, Dinoceras may be taken. This animal was but
little inferior to the elephant im bulk, but its brain capacity
was not more than one-eighth that of existing rhinoceroses.

The first Mesozoic Mammal in America was described by
Emmons, in 1857, from the Triassic of North Carolina.
Marsh, by his extensive discoveries, was enabled to fill up the
gaps to the Tertiary with many genera and species from the
western Jurassic and Cretaceous. Probably nine-tenths of all
the Mesozoic Mammals known in the world were described by
him, and while these remains are of great interest, yet from
their fragmentary condition they are not of the highest scien-
tific value, because little is known beyond the jaws and a few
limb bones.

In closing the outline of the discoveries made by this inves-
tigator, one cannot help being impressed with their signal
brilliancy, their great number, and especially by their unique
importance in the field of organic evolution. Were all other
evidence lost or wanting, the law of evolution would still have
a firm foundation in incontrovertible fact. The study of
variation and embryology in recent animals gives hints as to
the truth, but Paleontology alone can give the facts of descent.

Cuar_tres E. Brercuer.
YALE UNIVERSITY MUSEUM,
NEw HAVEN, Conn., May Ist, 1899.

Am. Jour. Sci.—FourtH Series, Vou. VII, No. 42.—Junzg, 1899.

420 Othniel Charles Marsh.

BIBLIOGRAPHY.

/

1861 The Gold of Nova Scotia. This Journal (2), vol. 32, pp. 395-400.
1862 On the Saurian Vertebre from Nova Scotia. Jdid., vol. 33, p. 278.
Description of the Remains of a new Enaliosaurian (Zosaurus Acadia-
nus), from the Coal Formation of Nova Scotia. /dzd., vol. 34, pp.
1-16, pls. i-ii.
1863 Catalogue of Mineral Localities in. New Brunswick, Nova Scotia,
and Newfoundland. Jdzd., vol. 35, pp. 210-218.

On the Science of the International Exhibition. /ézd., pp. 256-259.
1864 Notice of a new Fossil Annelid (Helminthodes antiguus) from the
Lithographic Slates of Solenhofen. Jdzd., vol. 38, p. 415.

1865 New genus of Jurassic Annelides (Istiynocemeen LZeitschr. deutsch.
ecolrGesc. »Vol,.W7, p.. 13, serine
Double Lobe-lines of Ceratites nodosus. Tbid., pp. sale
1866 Description of an Ancient Sepulchral Mound near Newark, Ohio.
This Journal (2), vol. 42, pp. I-II.
1867 Discovery of additional Mastodon remains at Cohoes, New ue
Tbid., vol. 43, pp. 115-116.
Notice of anew Genus of fossil Sponges from the Lower Silurian.
Lbid., vol. 44, p. 88.
Gcumriblions to the Mineralogy of Nova Scotia, No. 1. Ledererite
identical with Gmelinite. TLbid., pp. 362-367. | piety’.
1868 On the Paleotrochis of Emmons from North Carolina. /d7d., vol. 45,
pp. 217-219. 7
On the Origin of the so-called Lignzlites or Epsomites. Abstract :
Proc. Amer, Assoc, Adv. Sct., vol. 16, pp. 135-143.
On some New Fossil Sponges from the Lower Silurian. Abstract:
Lb... 301.
On certain Effects produced upon Fossils by Weathering. Abstract :
Lbid., Dp. 305.
Obvelrations on the Metamorphosis of Szredon into see ere
This Journal (2), vol. 46, pp. 364-374, 1 pl.
Notice of a new and diminutive Species of Fossil Horse (Zguus
parvulus), from the Tertiary of Nebraska. Jdzd., pp. 374-375.
1869 Notice of some New Reptilian Remains from the Cretaceous of Brazil.
Ibid., vol. 47, pp. 390-392.
Blesoripuion of a New and Gigantic Fossil Serpent (Dinophis grandis),
from the Tertiary of New Jersey. Jdzd., vol. 48, pp. 397-400.
Description of a New Species of Protichnites from the Potsdam Sand -
stone of New York. Proc. Amer. Assoc. Adv, Sci., vol. 17, pp.
322-324.
On the Preservation of Color in Fossils from Palzozoic eee A BOn
Lbid., pp. 325-326. ;
Ona ete neg Locality of Vertebrate Remains in the Tertiary of
Nebraska. Abstract: Canadian Naturalist, vol. 4, pp. 322-323.

Othniel Charles Marsh. 49)

1869 Notice of some new Mosasauroid Reptiles from the Greensand of

New Jersey. Zhis Journal (2), vol. 48, pp. 392-397.
Notice of some New Tertiary and Cretaceous Fishes. Proc. Amer.
Assoc. Adv. Sct., vol. 18, pp. 227-230.

1870 Notice of some Fossil Birds from the Cretaceous and Tertiary Forma-

1871

1872

tions of the United States. This Journal (2), vol. 49, pp. 205-217.

Note on the Remains of Fossil Birds. J/dzd., p. 272.

Notice of a new Species of Gavial from the Eocene of New Jersey.
LIbid., vol. 50, pp. 97-99.

Note on Lofhiodon from the Miocene of New Jersey. Proc. Acad.
Nat. Sct. Phila., vol. 23, pp. 9-10.

New Reptiles and Fishes from the Cretaceous and Tertiary Forma-
tions. Jbid., pp. 103-105.

On the Geology of the Eastern Uintah Mountains. TZhis Journal (3),
vol. I, pp. 191-198.

Notice of a Fossil Forest in the Tertiary of California. /d7zd., pp.
266-268.

Description of some new Fossil Serpents from the Tertiary Deposits
of Wyoming. Jdzd., pp. 322-329.

Notice of some New Fossil Reptiles from the Cretaceous and Terti-
ary Formations. J/d7d., pp. 447-459.

Note on a new and gigantic Species of Pterodactyle. Jézd., p. 472.

Notice of some new Fossil Mammals from the Tertiary Formation.
Tbid., vol. 2, pp. 35-44.

Notice of some new Fossil Mammals and Birds from the Tertiary
Formations of the West. Jd7d., pp. 120-127.

Discovery of a remarkable Fossil Bird. Jézd., vol. 3, pp. 56-57.

Explorations in Rocky Mountains, Oregon, etc. Abstract: Proc.
Cal. Acad. -Nat. Sct., vol. 4, p. 200.

Discovery of Additional Remains of Pterosauria, with descriptions of
two new Species. TZ%is Journal (3), vol. 3, pp. 241-248.

Discovery of the Dermal Scutes of Mosasauroid Reptiles. Jdzd., pp.
290-292.

Notice of a new Species of Hadrosaurus. Tbid., p. 301.

Preliminary Description of Hesperornis regalis, with Notices of four
other new Species of Cretaceous Birds. J/dzd., pp. 360-365.

On the Structure of the Skull and Limbs in Mosasauroid Reptiles,
with descriptions of new Genera and Species. Jdzd., pp. 448-464,
pls. x—xiii.

Boulders in Coal. Amer. Naturalist, vol. 6, p. 439.

Preliminary Description of new Tertiary Mammals. Pt. I. TZhis
Journal (3), vol. 4, pp. 122-128.

Note on Rhinosaurus. Tbid., p. 147.

Preliminary Descriptions of new Tertiary Mammals. Pts. II, III,
and IV. J/did., pp. 202-224.

Notice of some new Tertiary and Post-Tertiary Birds. /éid., pp. 256-
262. .

Preliminary Description of new Tertiary Reptiles. Pts. I and II.
lbid., pp. 298-309.

499, Othniel Charles Marsh.

a ee YS
—— $a

»

1872 Note on Zznoceras anceps. Tbid., p. 322.

Notice of a New Species of Zznoceras. Tbid., p. 323.

Notice of some Remarkable Fossil Mammals. J/dézd., pp. 343-344.

Notice of a New and Remarkable Fossil Bird. Jdzd., p. 344.

Discovery of Fossil Quadriumana in the Eocene of Wyoming. Jézd.,
Pp. 405-406. |

Note on a New Genus of Carnivores from the Tertiary of Wyoming.
Lbid., p. 406.

Notice of a New Reptile from the Cretaceous. J/dzd., p. 406.

Discovery of new Rocky Mountain fossils. Proc. Amer. Philos. Soc.,
vol. 12, pp. 578-579.

Synopsis of American Fossil Birds. Coues’s ‘‘ Key to North Ameri-
can Birds,” Salem, 8°, pp. 347-350.

1873 Notice of a new Species of /chthyornis. This Journal (3), vol. 5,

Pp. 74.

On some of Professor Cope’s Recent Investigations. Amer. Natural-
7St | VOlemgp pI. 51-52.

On the Gigantic Fossil Mammals of the Order Dinocerata. This
Journal (3), vol. 5, pp. 117-122, pls. i-ii.

On a New Sub-class of Fossil Birds (Odontornithes), Jbtd., pp. 161-

att

_

162.

Fossil Birds from the Cretaceous of North America. Jdid., pp. 229-
230.

Notes on the Dates of some of Prof. Cope’s recent Papers. Jdid.,
Pp. 235-236.

The Fossil Mammals of the Order Dinocerata. Amer. Naturahst,
vol. 7, pp. 146-153, pls. i-ii.

Additional Observations on the Dinocerata. This Journal (3), vol. 5,
Pp. 293-296.

Supplementary Note on the Dinocerata. Jbid., pp. 310-311.

On the Genus 7izoceras and its Allies. Amer. Naturalist, vol. 7, pp.
207-28).

Notice of New Tertiary Mammals. TZhis Journal (3), vol. 5, pp. 407-
410,

On the Dates of Prof. Cope’s Recent Publications. Amer. Naturakst,
vol. 7, pp. 303-306.

Tinoceras and its Allies. Jdzd., pp. 306-308.

Reply to Professor Cope’s Explanation. J/ézd., Appendix, pp. i-ix.

Notice of New Tertiary Mammals (continued). Zhzs Journal (3), vol.
5, pp. 485-488.

New Observations on the Dinocerata. Tbid., vol. 6, pp. 300-301.

On the Gigantic Mammals of the American Eocene. Proc. Amer.
Philos. Soc., vol. 13, pp. 255-256.

1874 On the Structure and Affinities of the Grontotheride. This Journal

(3), vol. 7, pp. 81-86, pls. i-ii.

Notice of New Equine Mammals from the Tertiary Formation.
Lbid., pp. 247-258.

Fossil Horses in America. Amer, Naturalist, vol. 8, pp. 288-294.

Othniel Charles Marsh. 493

1874 Notice of New Tertiary Mammals. Pt. III. 7hzs Journal (3), vol. 7,
PP. 531-534.
Small size of the Brain in Tertiary Mammals. Jdé7d., vol. 8, pp. 66-
67.
1875 Ancient Lake Basins of the Rocky Mountain Region. Pt. I. Jéid.,
vol. 9, pp. 49-52.
Results of Rocky Mountain Expedition. Abstract: J/ézd., p. 62.
New Order of Eocene Mammals. Jézd., p. 221.
Notice of New Tertiary Mammals. Pt. IV. Jdzd., pp. 239-250.
A Statement of Affairs at Red Cloud Agency, made to the President
of the United States. Rept. Special Commission to investigate Affairs
Red Cloud Indian Agency, pp. 1-113, Washington.
- Note on Reindeer Bones from a Clay Pit near North Haven. Tis
Journal (3), vol. 10, pp. 354-355.
On the O2ontornithes, or Birds with Teeth. Jdzd., pp. 403-408, pls.
ix—x.
1876 Principal Characters of the Dinocerata. Pt. I. Jbid., vol. 11, pp. 163-
168, pls. i—vi.
Principal Characters of the 77/lodontia. Pt. 1. Jbid., pp. 249-251,
pls. viii-ix.
Principal Characters of the Brontotheride, Tbid., pp. 335-340, pls. i-iv.
On some of the Characters of the genus Coryphodon, Owen. Jbid.,
pp. 425-428, 1 pl.
Notice of a new Sub-order of Péerosauria, Tbid., pp. 507-509.
Notice of new Odontornithes. Tbid., pp. 509-511.
Recent Discoveries of Extinct Animals. J/dzd., vol. 12, pp. 59-61.
Notice of New Tertiary Mammals. Pt. V. J/did., pp. 401-404.
Principal Characters of American Pterodactyls. /ézd., pp. 479-480.
1877 Brain of Coryphodon. Amer. Naturalist, vol, 11, p. 375.
Principal Characters of the Coryphodontide. This Journal (3), vol.
14, pp. 81-85, pl. iv.
Characters of the Odontornithes, with Notice of a new allied Genus.
Lbid., pp. 85-87, pl. v.
Notice of a new and Gigantic Dinosaur. J/did., pp. 87-88.
Notice of some new Vertebrate fossils. J/did., pp. 249-256.
Introduction and Succession of Vertebrate Life in America. Mature,
vol. 16, pp. 448-450, 470-472, and 489-491, London; and T7his
Journal (3), vol. 14, pp. 338-378.
A New Order of extinct Reptilia (Stegosauria\, from the Jurassic of
the Rocky Mountains. /d¢7d., pp. 513-514.
Notice of New Dinosaurian Reptiles from the Jurassic Formation.
Lbid., pp. 514-516.
1878 New Species of Ceratodus, from the Jurassic. Jéid., vol. 15, p. 76.
Notice of New Dinosaurian Reptiles. /¢id., pp. 241-244.
Notice of New Fossil Reptiles. J/éid., pp. 409-411.
Fossil Mammal from the Jurassic of the Rocky Mountains. Jdzd.,

Pp. 459.
New Pterodactyl from the Jurassic of the Rocky Mountains, /d7d.,

vol. 16, pp. 233-234.

494 Othniel Charles Mi ea

1878 Principal Characters of American Jurassic Dinosaurs. Pt. I. Jdid.,
pp. 411-416, pls. iv-x.
Scientific Museums. gth Ann. Rept. Amer. Mus. Nat. Hist., pp. 52-54.
1879 A new Order of Extinct Reptiles (Sawranodonta), from the Jurassic
Formation of the Rocky Mountains. TZhis Journal (3), vol. 17,
pp. 85-86.
Principal Characters of American Jurassic Dinosaurs. Pt. II. /ézd.,
pp. 86-02, pls. iii-x.
Additional Characters of the Sauropoda. Jbid., pp. 181-182.
The Vertebre of Recent Birds. Jdid., pp. 266-260.
Polydactyl Horses, Recent and Extinct. Jd7d., pp. 499-505, I pl.
Notice of a New Jurassic Mammal. J/dzd., vol. 18, pp. 60-61.
Additional Remains of Jurassic Mammals. J/d2d., pp. 215-216.
History and Methods of Palzontological Discovery. Mature, vol. 20,
PP. 494-499 and 515-521, London; and 7his Journal (3), vol. 18,
PP- 323-359.
Notice of New Jurassic Mammals. Jdzd., pp. 396-398.
Notice of New Jurassic Reptiles. J/dzd., pp. 501-505, pl. iii.
Peabody Museum. Vale Book, vol. 2, pp. 178-186.
1880 New Characters of Mosasauroid Reptiles. Zzs Journal (3), vol. 19,

pp. 83-87, pl. i. | ,
The Limbs of Sauranodon, with Notice of anew Species. J¢id., pp.
169-171.

Principal Characters of American Jurassic Dinosaurs. Pt.III. J/dzd.,
PP. 253-259, pls. vi-xi. ;

The Sternum in Dinosaurian Reptiles. Jdzd., pp. 395-396, pl. xviii.

Note on Sauranodon. Tbid., p. 491. : |

Odontornithes: a Monograph on the Extinct Toothed Birds of North
America. With 34 plates and 4o woodcuts. 4°, xv+20I pp. U.
S. Geol, Exploration goth Parallel, vol. 7, Washington ; and Wem.
Peabody Mus. Vale Coll., vol. I.

Notice of Jurassic Mammals representing two New Orders. This
Journal (3), vol. 20, pp. 235-239.

1881 Principal Characters of American Jurassic Dinosaurs. Pt. IV.
Spinal Cord, Pelvis, and Limbs of Stegosaurus. JTbid., vol. 21,
pp. 167-170, pls. vi-viii. .

A New Order of Extinct Jurassic Reptiles (Celuria), Jbid., pp. 339-
B40. plex.

Discovery of a Fossil Bird in the Jurassic of Wyoming. Jé7d., pp.
341-342.

Note on American Pterodactyls. Jdzd., pp. 342-343.

Principal Characters of American Jurassic Dinosaurs. Pt. V. Jdid., .

“pp. 416-423, pls. xii-xviii.
Notice of New Jurassic Mammals. Jdzd., pp. 511-513.
Restoration of Dinoceras mirabile. Tbid., vol. 22, pp. 31-32, pl. ii.
Jurassic Birds and their Allies. Sczence, vol. 2, pp. 512-513.
1882 Classification of the Dinosauria. This Journal (3), vol. 23, pp. 81-86.
The Wings of Pterodactyles. Jdzd., pp. 251-256, pl. iii.

ie owe ee

Othniel Charles Marsh. 425

1883 Evolution. ‘‘ Herbert mpences on the Americans and the Americans
on Herbert Spencer,” pp. 45-50.
Birds with Teeth. 37rd Ann. Rept. Director U. S. Geol, Surv., pp.
45-88.
Principal Characters of American Jurassic Dinosaurs. Pt. VI. Res-
toration of Brontosaurus. This Journal (3), vol. 26, pp. 81-85, pl. i.
On the supposed Human Footprints recently found in. Nevada.
Lbid., pp. 139-140.
1884 Principal Characters of American Jurassic Dinosaurs. Pt. VII.
On the Diplodocide, a New Family of the Sauropfoda. Tbid., vol.
27, pp. 161-167, pls. iii-iv.
Principal Characters of American Jurassic Dinosaurs. Pt. VIII.
The Order Theropoda. Tbid., pp. 329-340, pls. viii-xiv.
A New Order of extinct Jurassic Reptiles: ((/acelognatha). Jbid.,
p. 341.
Principal Characters of American Cretaceous Pterodactyls. Pt. I.
The Skull of Pteranodon. Tbid., pp. 423-426, pl. xv.
On the United Metatarsal Bones of Ceratosaurus. Tbid., vol. 28, pp.
161-162.
On the Classification and Affinities of Dinosaurian Reptiles. Mature,
- vol. 31, pp. 68-69.
1885 The Gigantic Mammals of the Order Dinocerata. 5th Ann. Rept.
Director U. S. Geol. Surv., pp. 243-302.
On American Jurassic Mammals. eft. Brit. Assoc. Adv, Sct. for 1884,
Pp. 734-736, London. |
Names of Extinct Reptiles. Zs Journal (3), vol. 29, p. 169.
On the Size of the Brain in Extinct Animals. Abstract: Mature, vol.
32, p. 562, London.
1886 Dinocerata: a Monograph of an Extinct Order of Gigantic Mammals.
| With 56 plates and 200 woodcuts. 4°, xviiit237 pp. Monographs
U.S. Geol, Surv., vol. to, Washington, (Author’s edition, title
page dated 1884; published 1885.)
1887 American JurassicMammals. This Journal (3), vol. 33, pp. 327-348,
pls. vii-x.
Notice of New Fossil Mammals. /J/dzd., vol. 34, pp. 323-331.
Principal Characters of American Jurassic Dinosaurs. Pt. IX. The
Skull and Dermal Armor of Stegosaurus. Tbid., pp. 413-417, pls.
vi-ix,
1888 Notice of a New Genus of. Sauropoda and other new Dinosaurs from
the Potomac Formation. /dzd., vol. 35, pp. 89-94.
Notice of a new Fossil Sirenian, from California. /ézd., pp. 94-096.
A New Family of Horned Dinosauria from the Cretaceous. J/did.,
vol. 36, pp. 447-478, pl. xi.
1889 Restoration of Brontops robustus from the Miocene of America.
Abstract: /dzd., vol. 37, pp. 163-165, pl. vi.
A Comparison of the Principal Forms of Dinosauria of Europe and
America. Abstract: /did., pp. 323-331.
Notice of new American Dinosauria. Tbid., pp. 331-336.

Othniel Charles Marsh.

eameee Characters of the order Hallopoda. Tbid., pp. 415-417.

Additional Characters of the Ceratopside, with notice of new Creta-
ceous Dinosaurs. Jd7zd., pp. 418-426, pls. v—vii.

Notice of New Tertiary Mammals. /dzd., pp. 523-525.

Notice of some extinct Zestudinata. Jbid., vol. 40, pp. 177-179, pls.
vii-viii.

1891 A Horned Artiodactyle (Protoceras celer) from the Miocene.

vol. 41, pp. 81-82.

On the Gigantic Ceratopside, or Horned Dinosaurs, of North America.
Lbid., pp. 167-178, pls. i-x.

On the Cretaceous Mammals of North America. Abstract:
Brit. Assoc. Adv. Sci. for 1890, pp. 853-854, London.

Restoration of 7receratops [and ee This Journal (3), vol.
41, pp. 339-342, pls. xv-xvi.

Note on Mesozoic Wammalia.

Restoration of Stegosaurus.
pl. 1x,

Notice of New Vertebrate Fossils. /ézd., pp. 265-269.

Geological Horizons as determined by Vertebrate Fossils.
336-338, pl. xii.

1892 The Skull of Zorosaurus.

Discovery of Cretaceous Mdnmiatie
pls. v—xi.

Recent [and extinct] Polydactyle Horses.

Ibid,

Rept.

Amer. Naturalist, vol, 25, pp. 61:-616.
This Journal (3), vol. 42, pp. 179-181,

Lbid., pp.

Pt; IIT. vo. pp. eieeeee.

Lbid., pp. 339-355.

A New Order of Extinct Eocene Mammals (Mesodactyla). Jbid., pp.
445-449.

Notice of New Reptiles from the Laramie Formation. J/dzd., pp.
449-453.

lbid., pp. 543-546, pls. xv—xvii.
Loid., voll "44, pp tga 7o,

Notes on Triassic Dinosauria.

Notes on Mesozoic Vertebrate Fossils.
pls. qi=v~

Restorations of Claosaurus and Ceratosaurus.
vi-vii.

Restoration of Mastodon Americanus, Cuvier. Jbid., p. 350, pl. viii.

1893 A New Cretaceous Bird allied to Hesperornis. Tbid., vol. 45, pp. 81-82.
The Skull and Brain of Claosaurus. Jbid., pp. 83-86, pls. iv—v.

Lbid., pp. 343-349, pls.

ah; Restoration of Azchisaurus. TLbid., pp. 169-170, pl. vi.
9 Restorations of Anchisaurus, Ceratosaurus, and Claosaurus. Geol. Mag.
ih (3), vol. x; pp, 15n—152, London?
ly |
ali
ria
fi
Lait

426
1889 Discovery of Cretaceous Mammalia. Tbid., vol. 38, pp. 81-92, pls.
ii-v. ee
Notice of Gigantic Horned Dinzosauria from the Cretaceous. Jdid.,
PP. 173-175.
Discovery of Cretaceous Mammaha. Pt. Il. J/did., pp. 177-180, pls.
vii-viii.
The Skull of the Gigantic Ceratopside. Abstract: J/bid., pp. 501-506,
pl. xu
1890 en eat fe of New Dinosaurian Reptiles. Jézd., vol. 39, pp. 81-86,
pl.

—

Othniel Charles Marsh. 42%

1893 Some Recent Restorations of Dinosaurs. ature, vol. 48, pp.
437-438, London.

Restoration of Coryphodon. This Journal (3), vol. 46, pp. 321-326,
pls. v—-vi.

Description of Miocene Mammalia. Jbid., pp. 407-412, pls. vii-x.

1894 Restoration of Camptosaurus. LIbid., vol. 47, pp. 245-246, pl. vi.

Restoration of Llotherium. Tbid., pp. 407-408, pl. ix.

A New Miocene Mammal. /J/dzd., p. 409.

Footprints of Vertebrates in the Coal Measures of Kansas. dzd., vol.

The Typical Orzzthopoda of the American Jurassic. This Journal (3),
vol. 48, pp. 85-90, pls. iv-vii.

Eastern Division of the Miohippus Beds, with Notes on some of the
Characteristic Fossils. J/dzd., pp. 91-94.

Miocene Artiodactyles from the Eastern Miohippus Beds. Jdézd., pp. .
175-178.

Description of Tertiary Artiodactyles. Jdid., pp. 259-274.

A Gigantic Bird from the Eocene of New Jersey. Jdzd., p. 344.

A New Miocene Tapir. Jdzd., p. 348.

1895 On the Pithecanthropus erectus, Dubois, from Java. Jdzd., vol. 49, pp.
144-147, pl. ii.

Thomas Henry Huxley. J/d7d., vol. 50, pp. 177-183.

The Aepftiia of the Baptanodon Beds. Jdzd., pp. 405-406.

Restorations of some European Dinosaurs, with Suggestions as to
their Place among the Reftiia. Abstract: Jbid., pp. 407-412,
pls. v—viii.

On the Affinities and Classification of the Dinosaurian Reptiles.
Abstract: Jbdzd., pp. 483-498, pl. x. Reprinted with alterations,
under the title ‘‘ Classification of Dinosaurs.” Geol. Mag. (4),
vol. 3, pp. 388-400, London.

Fossil Vertebrates. Johnson's Universal Cyclopedia, newed., vol. 8,
Pp. 491-498, I pl.

Note on Globular Lightning. Mature, vol. 53, p. 152, London.

1896 The Age of the Wealden. This Journal (4), vol. 1, p. 234.

On the Pithecanthropus erectus, from the Tertiary of Java. Abstract:
Bid... .Pp> 475-482, pl. ‘xiii.’ Reprnted under the ‘title’ "The
Apeman from the Tertiary of Java.” Science, vol. 3, pp. 789-793.

A new Belodont Reptile (Stegomus) from the Connecticut River Sand-
stone. This Journal (4), vol. 2, pp. 59-62, pl. i.

The Geology of Block Island. Jézd., pp. 295-298.

Amphibian Footprints from the Devonian. J/did., pp. 374-375.

The Geology of Block Island (continued). /é%d., pp. 375-377.

The Dinosaurs of North America. 76th Ann. Rept. Director U. S.
Geol, Surv., Pt. I, pp. 133-414, pls. ii-lxxxv.

The Jurassic Formation on the Atlantic Coast. Abstract: This
Journal (4), vol. 2, pp. 433-447.

Vertebrate Fossils [ of the Denver Basin]. Monographs U. S. Geol.
Surv., vol. 27, pp. 473-550, pls. xxi-xxxi, Washington.

498 Pre Ohne! Charles Marsh.

1897 The Stylincdontia, a Suborder of Eocene Edentates. This Journal —
(4), vol. 3, pp. 137-146. | a
The Affinities of Hesperornis. Tbid., pp. 347-348. ‘
Principal Characters of the Protoceratide. Tbid., vol. 4, pp. 165-176,
pls. ii-viii. 3
The Skull of Protoceras. Geol. Mag. (4), vol. 4, pp. 433-439, pl.
xix, London. fou
Recent Observations on European Dinosaurs. This Journal (4), vol.
4, Pp. 413-416.
1898 New Species of Ceratopsia. Tbid., vol. 6, p. 92.
The Jurassic Formation on the Atlantic Coast. —Supplement. _Lbid.,
pp. 105-115. :
Cycad Horizons in the Rocky Mountain meager Lbid., pi iGz.”
The Value of Type Specimens and Importance of their Preservation.
Ibid., pp. 401-405.
The Onan of Mammals. Jdzd., pp. 406-409.
The Comparative Value of Different Kinds of Fossils in DeteLanaine =
Geological Age. Abstract: Jdzd., pp. 483-486.
On the Families of Sauropodous Dinosauria, Abstract: Jbid., pp.
487-488. .
1899 Footprints of Jurassic Dinosaurs. Jdid., vol. 7, pp. 227-232, pl. Wii
Note on a Bridger Eocene Carnivore. iia ps GOn

ss
eed

>
=”

- . z 2S Pee . : - a = = S eS
——— y Pa ES SF cian. a> =—- _= =a Stk = = se : = = =

|

; es *
a
ie

Safford and Schuchert—Camden Chert of Tennessee. 429

Art. XLVI.—The Camden Chert of Tennessee and its Lower
Oriskany Fauna; by JAMES M. SAFFORD and CHARLES
SCHUCHERT.*

Part I. The Camden Chert.
By JAMES M. SArrorD, State Geologist.

In March, 1855, the writer discovered in Benton County,

Tennessee, at several points, excellent outcrops of -Lower

Helderberg shales and limestones, very rich in fossils. The
discovery was important, since it settled the question as to the
presence of the Lower Helderberg, as a distinct formation, in
Tennessee, west of the meridian of Nashville. In subsequent
years, this discovery also led to a recognition of the chert now
referred to as Oriskany, which I have designated the Camden

chert for the reason that at Camden, the county-seat of Benton,

is seen one of its best exposures. |

One of the localities discovered in Benton County was a
bluff on Big Sandy River, about five miles from its mouth, at
a point then in Henry County, and known as the old William’s
Mill Site. This locality is referred to in “ Geology of Ten-
nessee,” 1869. As stated on page 325 of that book, there are
here exposed about 50 feet of bluish limestone, mostly shaly.
Above this and running back on a slope from the precipitous
portion of the exposure, ‘are loose, angular, flinty masses, con-
taining the fossils of the rocks below, and derived from cherty
layers not seen.” The fossils in the chert were not numerous
nor in good condition, but what was seen of them led, at the
time, to the conclusion stated.

In 1884, I recognized the chert at Camden as a distinct for-
mation. JI had, in passing through the country, seen this
horizon and had referred it without special examination to the
“Silicious Group ” (Lowest of Subcarboniferous), outcrops of
which, very like the chert of Camden, are seen at many points
in Benton and counties north and south of it. In my excur-
sion of 1884, however, I stopped for some time at Camden to
study the formations. The fossils in the chert arrested my
attention, and reminded me of those in the flints seen at
William’s Mill in 1855. But at Camden, the chert was in com-
paratively great force, at least 60 feet of it being exposed. At
first, I was inclined to consider the chert a division of the
Lower Helderberg, but subsequent studies of the fossils at
home forced me to the conviction that, as a group, they must

* Published with the permission of the Secretary of the Smithsonian Institu-
tion.

430 Safford and Schuchert—Camden Chert of T on nessee

be Oriskany. The fact that the formation was one of chert
also pointed to this.

Afterwards, in 1885, 1886, and 1887, I visited localities
where I thought the Camden might outcrop. One of these is
Big Sandy Station, in Benton, on the Memphis branch of the
Lonisville & Nashville Railroad, and near the point where the
road crosses Big Sandy River. Here I found the chert well
developed and abounding in fossils. The outcrops are as
extensive and as good as at Camden. For several miles south
of Big Sandy, the chert appears on the hillsides as loose angu- |
lar gravel.

ee miles arb. on the Lower Camden road, Lower Helder-
berg limestones are seen cropping out from beneath Camden
chert, with Tertiary beds also overlapping all in unconformable
contact.

In Henry County the Camden chert outcrops in considerable
areas, west and south of the William’s Mill locality. it is seen
in limited thickness above the Lower Helderberg in Decatur
County, and in the same relation, east of the Tennessee River,
in Stewart County. In the latter locality, it outcrops in the
bluff on the Cumberland River below Cumberland City. The
greatest development of it, however, is on the west side of the
Tennessee River, in a strip of country lying in Henry, Benton,
and Decatur Counties.

In 1897, I called the attention of Mr. Schuchert to the
Camden chert, at the same time trusting he might be able to
visit the Camden locality. This he did, collecting a series of
fossils, which he studied, kindly giving me the results. Jam
under special obligation to him for this visit.

Part II. The Camden Lower Oriskany. |

By CHARLES SCHUCHERT.

In the spring of 1897, the writer collected for the U.S.
National Museum Lower Helderberg fossils in western Ten- »
nessee, and while in Nashville, Professor Safford also directed
his attention to a lot of Camden chert organisms. Since no
strata of Oriskany age had been recorded in Tennessee, the
importance of determining the equivalency of the Camden
chert with other regions made it desirable to know more of its
fauna, and with that object in view, a collection was made at
Camden.

The fossils of the Camden chert are, as a rule, natural casts
of both the interior and exterior of the organism, and preserve
the fined markings in detail. This fauna is closely related to
that described by Meek and Worthen* from the “Clear Guess

* Geol. Surv., IIT, vols. i, ii, and iti.

and its Lower Oriskany Fauna. 431

Limestone” of southern Lllinois, Alexander, Jackson, and
Union Counties. From this region are known but 11 species,
and 8 of these are also foundin Tennessee. They are Anoplia
nucleata, Anoplotheca flabellites, Hatonia peculiaris, Spirifer
worthenanus, S. hemicyclus, Megalanteris condoni, Amphi-
genia curta and Strophostylus cancellatus.

The “ Clear Creek limestone” of Illinois is intimately con-
nected with the Lower Helderberg below, and is not less than
200 feet thick, being followed by a “‘quartzose sandstone ”
from 40 to 60 feet in thickness. The latter is probably equiva-
lent to the Upper, or typical, Oriskany of New York, and
does not appear to be present in western Tennessee. From
Professor Safford’s description of the Camden chert, it is evi-
dent that the Lower Oriskany thins rapidly southward. In
Tennessee, it is about 60 feet in thickness, while it is not less
than 200 feet thick in Illinois exclusive of the Upper Oriskany,
which is entirely absent in the former state.

The “Camden chert” fauna contains 32 species, and 6
of these are restricted to southern Illinois and western Ten-
nessee. Of the entire fauna, 24 species are found either in
the Lower Helderberg or in the Lower Oriskany of other
regions, and 20 occur in the Upper Oriskany or Corniferous.
After removing the 13 species common to both the Lower and
Upper Oriskany, and the two restricted forms, 17 remain. Of
these 10 occur either in Lower Helderberg of Lower Oriskany
rocks of other regions, while 6 are found in higher beds. This
evidence, therefore, indicates clearly a Lower Oriskany age for
the “Camden chert” of Tennessee and the ‘Clear Creek
limestone” of Illinois, which indication is the more marked
because of the absence here of such characteristic Upper
Oriskany species as Lipparionyx proximus, Chonostrophia
complanata, Spirifer arenosus, Rensseleria ovoides, Meristella
lata, Camarotachia pleiopleura, C. barrandet, or C. speciosus.

Lower Oriskany fauna of the Camden Chert of Camden, Benton
County, Tennessee.
In Lower In Upper
Oriskany Oriskany
elsewhere. elsewhere.

Zaphrentis reemeri Hall? ._--..-...-2225_- Pe i*
mwondops terminalis: Hall... 2.2 2c222252 4 x
Hipparionyx proximus Vanuxem? A very
small specimen if this species_--------- x x
Chonetes mucronatus Hall? ............- x
Sd melonica Billings 4.<,))44-2<6-28 x

x
Chonostrophia reversa (Whitfield) ..-... - U. Ey

*L. H. = Lower Helderberg; U. H. = Upper Helderberg. This list is based
on the collection in the U. 8. National Museum. .

a

: ett Z. = :
= a on — = = =
ee —e ae +s es. SS ee

—
ae —_—

>
=

. -
t : , ae
oe 3

a
=

=
=

2

~—¥

a
=

a

eS.

SS"

432 Safford and Schuchert—Camden Chert of Tennessee.

Stropheodonta (Lepstrophia) perplana (Con-
Tad) ee teeeee Sk o ee
Orthothetes woolworthanus Hall.______.-
Anopliainucteata Halk. 2. 222i 22 Bee
Cyrtina affinis Billings..-.---.------.-.-
Metaplasia pyxidata Hall -.-..----.-.__-
Spirifer hemicyclus Meek and Worthen. .-
F prilottisGitall 2 ud US ee

f worthenanus Schuchert ___._._-_-
Anoplotheca flabellites (Conrad) -._._----
Meristella levis (Vanuxem) .-----_-..._-
yom) Et cv 6 (te page epee ok ORES BOER TS TE 2

Atrypa reticularis Linne ..-- 2...) 1_ <2
Katonia pecularis (Conrad) .--.----_-----
Amphigenia curta Meek and Worthen ?_--
Rensseleeria ovoides (Haton) ? _.-.._--__-
Megalanteris condoni (McChesney) ---- .-
Pentaeulites acwla Tall <0yk Sel ele
Anicula chevopscura Tiall =: ite a Aire
Actinopteria cfr. textilis Hall... -.------
Acrocoulia magnifica (Hall) ? ---..--..---
Acrocoulia (Orthonychia) tortuosa (Hall) ?
Diaphorostoma turbinata (Hall) -.--.---.
Strophostylus (?) cancellatus Meek and

Weorthemy fait 30 dea as Be
Phacops. cristata Halls... 230.2 4.50eee2
Ostracoda.

Total 32 species .--. - cee tee»

KK KX KX KOS

x x XX XXX xX

x

Whiteaves— Trenton. rocks at Ongava. 433

Art. XLVII.—Lecent discovery of rocks of the age of the
_ Trenton formation at Akpatok Island, Ungava Bay,
Ungava; by J. F. WHITEAVES.

WHILE accompanying the Hudson Bay expedition despatched
by the Canadian government in 1897, in the sealing steamer
“Diana,” Dr. R. Bell, of the Geological Survey of Canada,
spent the 13th of September on Akpatok Island, on his way
from Ashe Inlet to Fort Chimo. Ina preliminary report on
his explorations for that year,* Dr. Bell says: ‘The portion of
the island which I saw (from the northern end to the middle
of the east side) consists of unaltered gray limestone in hori-
zontal beds, and it presents a perpendicular wall 400 or 500
feet high all along.” ‘This sea-wall is clean-cut and the beds
appear thick and solid, but wherever their edges have been
long exposed to the weather, or in the hill-sides and ravines of
the interior, they split up into thinner layers.” ‘“ Some frag-
ments observed in one place had the appearance of lithographic
stone.” ‘I was enabled to land opposite the place where the
Diana anchored, as already mentioned, about the middle of
the eastern side, and I improved the opportunity to collect
fossils, which, however, were not abundant.” ‘ Those obtained
indicate the Hudson River formation.” :

When this last sentence was written, the specimens that it
refers to had not been unpacked. Since the publication of the
Report from which it and the four preceding sentences are
quoted (in May, 1898), these fossils have been examined by the
writer, and it at once became obvious, first, that they indicate a
little lower geological horizon than the Hudson River forma-
tion, viz., that of the Trenton limestone ; and, secondly, that
they are remarkably similar to the fossils of the Trenton
formation of the Red River valley in Manitoba. The collec-
tion consists of fifteen species, but two of these are imper-
fect casts of the interior of shells of gasteropoda, that can
only be determined generically. Of the thirteen that remain
eleven had previously been found in the Manitoba Trenton,
and nine are species that are common at East Selkirk and
Lower Fort Garry. The two fossils of which by far the most
specimens were collected are Streptelasma robustum and Cyrto-
ceras Manitobense. The former is a rather large rugose coral,
the types of which are from the Red River valley in Manitoba.

The following is a list of the species represented in the col-
lection, as far as they can be determined.

_ * Summary Report of the Geological Survey of Canada for 1897, page 82.

434 Whiteaves—Trenton rocks at Ungava. Ss

LIST OF FOSSILS FROM THE TRENTON FORMATION AT AKPATOK ISLAND,
COLLECTED BY DR. BELL IN 1897.

RECEPTACULITID®,
Receptaculites Oweni Hall.
A very small and worn but fairly characteristic fragment.

ei:

ZOANTHARIA.
' Streptelasma robustum W hiteaves.
About thirty specimens, of all sizes.

Calapecia Canadensis Billings.
One specimen.
BRACHIOPODA.
Rafinesquina lata Whiteaves.
One ventral valve.

Lepteena unicostata Meek and Worthen.
A tolerably well preserved ventral vaive that is apparently
referable to this species.

Plectambonites sericea (Sowerby).
A few specimens.

Orthis tricenaria Conrad.
An unusually large ventral valve.

Orthis (Dinorthis) Meedsi, var. arctica Schuchert.
Three imperfect and badly preserved specimens which Mr.
-Schuchert thinks look very much like Baftin Land specimens
that he is describing under this name.

Orthis (Hebertella) bellarugosa Conrad.
One good specimen that has been identitied with this species
by Mr. Schuchert.

Orthis (Dalmanella) testudinaria Dalman.
A few imperfect specimens.

Platystrophia biforata (Schlotheim).
Two imperfect ventral valves.

? Rhynchotrema ineequivalvis (Castelneau).
An imperfect ventral valve, with the cardinal area not visible.

GASTEROPODA.
Trochonema or Pleurotomaria. (Species indeterminable.)
A cast of the interior of the last volution and last but one.
Hormotoma. (Species indeterminable, but apparently like Z.
Saltert Ulrich.)

A east of the interior of three of the Bras.

~ = - - - — — — — > — ’ - : eae psn re aT T - =— = =
= - = =~ a : - 1} $s % - =: = 4 = = =
— — — - ———— = 4 = A ————<e = eo <————— ene - — > et a mend ~~ 4 pens. _——— > -— © e 9a
_ on = SS, ae ee cee — i“ = ~ = — a Sod = ek = - 5 = wenn = = ——. A a _ =
ees a SR a ee a a : = = = ae Saenis > a a ae Sees Toe Ts . z
= — ae > = eo a ss Saws a - oaty SS = = SER SS : —— = === - =— :
oer == S Sanna - — Se poe a - ————— —— - SSS - —

——S =

CEPHALOPODA.
Cyrtoceras Manitobense Whiteaves.
About thirty well preserved but much worn fragments, from
less than an inch to nearly two inches in length.
Ottawa, March 23d, 1899.

=

——-
SS

i 3) 3
— > a
Es -

rm Rl aS
ar

2 & t.. 3
Se

:
£
;
i

T. Holm—Studies in the Cyperacee. 435

Art. XLVIII.— Studies in the Cyperacee ; by THEo. Hou.
X. Fimbristylis Vahl; an anatomical treatise of North
American species. With fourteen figures in text, drawn
from nature by the author.

THAT purely floral characters are insufficient for the estab-

lishment of genera has been proved by several botanists in
_. recent years. This is especially the case with large orders

where a number of species are to be classified so as to demon-
strate, at least, a supposed relationship. It is not uncommon
that herbalists content themselves by studying only the floral
structure and matching the details with some old illustration
or analytical key; but there is, however, more than that to be
considered and looked upon, as affinity and relationship. The
establishment of natural orders is, of course, a progress, but, to
say the least, the genera themselves are often so badly under-
stood and so arranged, that we obtain no more knowledge of
the real affinities than if we had followed the system of Lin-
neeus: Monandria, Diandria and so on. It is very common to
see species arranged in genera widely apart, on account of the
stamens and stigmata being different in number, while the
plants otherwise show the same habits and internal structure.

Our genus Fimbristylis furnishes a good illustration. Merely

because some species have two or three stigmata, this genus has
been divided into Dichelostylis and Trichelostylis ; and the
fact that the base of the style is persistent in some species, but
only for some time, has caused the separation of the genera
Isolepis and Oncost, ylis.* Furthermore these genera have
been considered as subgenera of Sczrpus or even as close allies
of Heleocharis, a consideration, however, that cannot possibly
have been based upon observations in the field. But to
Bentham and Asa Gray such characters were of less importance,
and these authors were liberal enough to include a number of
species as true /imbristyles, even if the floral characters did
not come exactly within the scope of the original diagnosis or
conception of the genus as understood by Vahl. Finally the
anatomical study of a number of genera of the Cyperacew has
shown us that certain analogies exist which must be con-
sidered as valuable to the study of mutual aifinities, promul-
gated by Schwendener, Pallat and Rikli.

* Tt was Martius (Flor. Brasil.) who substituted the generic name Oncostylis for
Bulbostylis, which appears in the Conspectus generum, because DeCandolle had
already used Bulbostylis to designate a genus of Composite.

+ Palla Ed. Zur Kenntniss der Gattung “Scirpus.” (Engler’s botan. Jahr-
biicher, vol. x, Leipzig, 1889, p. 293.)

Am. Jour. Sci1.—FourtH Series, Vou. VII, No. 42.—Junz, 1899.

29

436 T. Holm—Studies in the Oyperacee.

Considering the genus Fimbristylis in North America not
only “‘sensu strictiori,” but inclading the species which Torrey

reterred to Isolepis : stenophylla, cilratifolia, Warec and

capularis R. et 8., we cannot avoid noticing that the general
habit of these plants, both annuals and perennials, is very
much the same. A dense cespitose growth is characteristic of
most of our species; the leaves are generally narrow with
short sheaths, and the flowers are arranged in true spikes,
either borne on long peduncles constituting open cymes, or
sessile in small heads. While a pubescence is rarely observed
in Cyperaceew, it is not uncommon in Himbristylis, of which
several species have the leaves, stems, involucres and scales
clothed with short hairs. In examining the inflorescence, the
spikes are mostly, as stated above, arranged in open cymes,
where the terminal spike is almost sessile and overtopped by
the lateral ones. In / castanea and./. thermalis several
lateral spikes are borne on long peduncles, but with no further
ramifications, representing respectively a di- or pleio-chasinm.
A similarly constructed cyme is, also, characteristic of F. laxa
in plants growing in the vicinity of Brookland, D. C.; in

specimens from Eustis, subtropical Florida, the lateral ves
cences are, on the other hand, decompound, thus agreeing with
vib spadicea. The pleiochasium i in /. puberula is either simple

or decompound ; in the first case the spikes are considerably

larger than in the last, where the spikes consequently are more
numerous. The most decompound inflorescence is, however,

to be found in /! autumnalis, where the secondary branches |

are pleiochasia, the tertiary, on ‘the contrary, passing over into
monochasia of only two spikes, a terminal and a lateral. While
~in # capillarcs the inflorescence is typically a pleiochasium, it
is not unusual to meet with specimens where the flowers are
reduced to asingle spike, at the base of which a long involueral
bract is often noticeable, but without supporting any lateral
branch. Similar empty involucral leaves, but reduced to seale-
like, long-pointed bracts, occur also at the bases of several of
the spikes, indicating the ‘place of non-developed inflorescential
branches. In species with sessile spikes, #. Warei and £.

stenophylla, the composition of the inflorescence becomes
more indistinct, since all the peduncles are very short and the
characteristic clado-prophyllon entirely suppressed, a fact that
is known also from other genera of Cyperacee with similarly
averegated spikes : species of CO; yperus, Dichromena, Scirpus,
Carex and others. But where a clado-prophyllon is developed,

this is readily noticed at the very base of each lateral peduncle
as a tubular sheath, more or less puberulent in our species of
Fimbristylis. It varies somewhat in shape, being compressed
in ££. autwmnalis and its more southern ally Fe complanata,

T. Holm—Studies in the Cyperacee. 437

or bicarinate in /. castanea and F’. spadicea, or it is cylindric
as in the other species; the apex is prolonged into two subn-
late teeth in F. laxa and F. castanea, but merely slightly
oblique in the others. While the inflorescence of F imbristylis
does not show any remarkable deviation from that of other
Cyperaceew, we have observed in the disposition of the leaves
a very peculiar exception from the ordinary rule in this order,
at least in some species. As invariably quoted in botanical
manuals, the tristichous leaf-arrangement is the characteristic
one of the C; yperacee, while on the contrary the distichous
that of the Graminew. It would hardly be natural to suppose
that any such rule should be stable for either of these orders,
since exceptions are known if not from the leaves then at least
from the arrangement of the spikelets and the inflorescential
branches. We might merely call attention to the panicles of a
number of Graminew, of- which the branches are not distich-
ous; besides that the spikelets are sometimes spirally arranged,

.as for instance in species of “ragrostis. In the Cyper ace

the distichous arrangement of the flowers with their bracts is

well known as the very characteristic of Cyperus, Dulichium,

partly also of Scleria and Dichromena. But ae. arrangement
of the stem-leaves has, so far, usually been described as tris-
tichous in the Cyperacee, in spite of the fact that Kunth, in

his diagnosis ot /. autumnalis, characterized the leaves as

“ distichis,” a statement which is absolutely correct. The
leaves of this species have laterally compressed sheaths, which
soon become split, and the blades are furthermore held ina
very peculiar position: with the one margin turned towards.
the axis or the flower-bearing stem. Although the leaf-blades

- are not isolateral, they invariably show this position, a fact that

we have never noticed before in any of our American species
with corresponding anatomical structure excepting in /. com-
planata. The distichous arrangement of the leaves is, how-
ever, also to be seen in /. castanea and F. spadicea, but the
leaf-blades are not turned to the side in these species; they
occupy the usnal position as dorsiventral leaves. Considered
from a morphological view-point, it seems strange that such
disposition of leaves should be found in a genus like #%mbris-
tylis, in which all the other foliar organs, involueral-leaves and
floral bracts are arranged in spirals; it would seem more natu-
ral to a genus like Cyperus or Dulichium with the distichous
flowers. And by examining the internal structure of such
leaves, it seems still more surprising that we find a number of
analogies rather than important variations. Furthermore when
we compare our North American species with each other, it
will be seen from the following pages, that even if they inhabit
localities differing greatly in climate and soil, their internal

438 T. Holm—Studies in the Cyperacec.

structure does not seem to be influenced to such an extent as
to render it impossible to detect several and very gradual tran-
sitions. The modern, but often too superficial and hasty
classification of plants as Xero- and Hydro-phytes, is not by
any means applicable to our genus, at least not to its North
American representatives. If such types of plants, Xero- and
Hydro-phytes, really exist, it must be proved not by observa-
tions in the field alone, but by a very close study of their struc-
tural peculiarities. The mere fact that such and such species 7
inhabit dry, sandy places or marshes, ponds, etc., does not
necessitate the plants in question to be considered as either |
xero- or hydro-philous species; their internal structure must
be carefully examined before any conclusion can be drawn, as
to whether these plants are “solely adapted” to these special |
environments. As regards our species of Himbristylis it
might seem very tempting to classify these in this way: /.
capularis, Ff. Warei, FE. stenophylla and F. ciliatifolia as
xerophytes, and the remaining species as hydrophytes, but
their anatomical structure does not seem to justify any such
classification. In order to facilitate the anatomical comparison |
of the material, which we have examined, it may be well to
enumerate the various localities as follows: /” autumnalis was
collected in wet sandy or clayish soil in the vicinity of
Washington, D. C., and in low sandy places near Eustis, sub-
tropical Florida; /. castanea was from brackish marshes near
the coast of Maryland; /. spadicea and F. puberula both from
wet, sandy soil near Eustis, Florida; /: Zawa from low meadow-
land, rather dry soil, near Brookland. D. C., as well as from
low, sandy soil near Eustis, Florida; 7 thermalis from hot
sulphur springs near Golkonda, Nevada; /. capillaris from -
dry, sandy soil near Washington, D. C., and Eustis, Florida;
Lf. ciliatifolia from dry, sandy fields in Alabama, and finally
LI”. stenophylla and F. Ware from dry, sandy soil in the High- |
Pine-woods near Eustis, Florida. Besides these we have, also, :
examined some specimens of /” complanata and F’. capillaris |
from Montevideo, Uruguay, which together with a number of
other Cyperacew were presented to the writer through the
courtesy of Professor Arechavaleta, Director of the National
Museum at Montevideo. :
In discussing the anatomy of the vegetative organs of these
species, we will begin with ;
| |

The leaf.

By its greater ability to vary, the leaf appears to illustrate
the species better than the stem or root. Considering the
mere outline of the leaf, three types are readily distinguished :

4 th

a

ie de i ia

a

oe

Ss

ri
y

T. Holm—sStudies in the Cyperacee. 439

(1) the relatively broad and flat, as we find it in /. autumnalis,
Ff’, thermalis, F. spadicea, F- puberula, F. complanata and
fF’. laxa ; (2) the semi-cylindrical in /. castanea and (8) the
approximately triangular in the remaining species. Very
characteristic of the leaves of the first category is the more or
less, but always very conspicuous, asymmetric blade, in contradis-
tinction to those ‘of the other species, a fact that has , also,
been recorded by Rikli as common to 7: glomerata (N ees) and
Ff’, paupercula (Bklr.) In -. autumnalis the leaf, as will be
seen from the accompanying illustration (fig. 1), is flat and
strictly dorsiventral; only the midrib is somewhat projecting
on the lower surface of the blade, and this is due to its larger
support of stereome. The upper face is perfectly smooth,

&
ae us
po

Maa: eee
Fic. 1. Transverse section of half of the leaf-blade of F autwmnalis. x5.

and there is nothing in the outline to indicate the peculiar
position of the blade, as we have mentioned already, being
placed edgewise by the twisting of the leaf-sheath. This form
of leaf must not be confounded with the sword-shaped leaves
of Lridacew, because both faces are equally well developed in
Fimbristylis autumnalis, besides in F. complanata.

The leaf is more likely to be compared with that of a mature
Eucalyptus or of our evergreen species of Smilax, of which
the leaves show a similar position. It is strange to see that
such peculiar leaf-position is said to be either “a xerophy tic
character of bog-plants or a natural adaptation of desert-plants.”
However in these species of /imbristylis the leaf-blade is
developed with this position immediately after germination,
and we have not noticed any exceptions to this rule in speci-
mens from less dry situations, nor in plants growing in
Sphagnum-moss. It is a character that seems to be peculiar to
these two species, but which cannot possibly be considered as
having originated through any means of adaptation, since the
structure is identical with that of the other species, in which
the ventral leaf-face is turned towards the stem.

While the epidermis of the upper face in /. autumnalis
consists only of one layer of large, thin-walled cells of equal
width all through the blade, we find in F. complanata several
layers, representing a ty pical water-storing tissue. Among the

440 T. Holm—Studies in the Cyperacee.

other species /” lawa (fig. 2) comes nearest to /. autumnalis,

since the superior epidermis is only composed of one layer,
but the leaf is much narrower and the blade is not placed
edgewise; furthermore the leaves are tristichous. In /%
spadicea the accompanying figure (fig. 3) shows that the epi-
dermis has been developed in a few strata, but only in certain
places, between the mestome-bundles. It is very strange that
this species has often been considered as identical with /.
castanea or at least as the type, since both are so very different,

Se

Fig. 2. Transverse section of leaf-blade of F. lawa. x75.

-Fig. 3. Transverse section of half of the leaf-blade of / spadicea. x15.

when considered from an anatomical point of view. By
examining the leaf of /. puberula (fig. 4) we notice that the
upper surface is occupied by a huge mass of colorless and rela-
tive large cells, which represent an epidermis of several layers.
Besides this the epidermis of the inferior surface is developed
as a dense covering of long, unicellular hairs, a fact that would

GaGuorsal <5

a
ee.

Fie. 4. Transverse section of half of the leaf-blade of F. puberula. x15.

seem to point towards some special adaptation, ‘‘a xerophytic
character developed in a hydrophilous plant”; yet the species
is neither a hydrophyte or xerophyte as these terms are gen-

T. Holm—Studies in the Cyperacee. 441

erally applied. A structure similar to that of /. puberula but
without hairs is to be found in / thermalis (fig. 5), of which
the writer received some excellent material from Dr. E. L.
Greene. The leaf-blade is here broad and very thick from the
immense layers of colorless tissue on the upper surface, the
function of which cannot possibly be for the storage of water,
since the specimens were growing in hot sulphur springs. /
castanea exhibits a very peculiar shape of leaf, this being semi-
cylindrical (fig. 6), and of which the upper surface shows a

Fic. 5. Transverse section of half of the leaf-blade of F&. thermdlis. x

still larger development of colorless tissue than any of the
other species we have examined. it was rather surprising to
find this narrow leaf developed on a species that inhabits
brackish marshes, which are not liable to become dried out
during the summer. It is actually the narrowest leaf we
have observed in any of the
North American species of
Fimbristylis, in proportion
to its length, and the only
species that possess approxi-
mately as narrow leaves are
just those which inhabit a
very dry and sandy soil:
fF’ capillaris, F. ciliatifolia,
LF. Warei aud F. stenophylla.
The accompanying figures
(figs. 7, 8 and 9) illustrate
transverse sections of leaves
from the latter three species, Fic. 6. Transvers
and we notice at once the of ¥ castanea. x
very narrow, almost triangular outline from the strongly pro-
jecting midrib. No large tissue for water-storage is developed
in any of these species, but the cells of epidermis are on both
surfaces relatively wide, and sometimes a double epidermis
may be developed between the mestome-bundles, in both /.
ciliatifolia and F. Warei. Characteristic of the leaf of F.

os . ; : . , - 5 : = i ' . — hs - » a ws : 7» ‘ . = ‘ 3 —— < =
-+ ~ : 7 = = 2 4 . ™~- iY . . we | om. 7 : r
~ . La - - 7 3 ¥ x . es . ah = . ‘= - - a
a : < . = 2 — == weet ee ———_—<——__— a —— = ——o — ;
waa ‘ ' e a : > mi <2 Si ate . es > en + Se :
NE SY ee meet. he BEE | =? SSA BLES “sD.
a | ee eye = — ~~ —>— — <7 7. —. Ya a ¥ > % = AM < "vn ina dain, - _ .
4 = - — - = Poa = .. - ~ = 4S > £ =
- : = i. ” Fy Ss — —— - = _ => ee i
= : — —- = —— ae —— > . “ = S13 : . ee ee: aa ks : : —r 2 % 32-2. —_— ¥
= —a a ee = += ee =———=S rR Ok ee 3 ee Se SS SS eee = a eee, Se = — = ——

449 T. Holm—Studies in the Cyperacee.

stenophylla (fig. 9) are the double rows of diverging, pointed ~
hairs along the margins and the ribs on the lower surface of
the blade.
Although /. complanata is not a native of this country, we
have examined some specimens, as stated above, from Uruguay,
which were originally labeled /’ autumnalis, to which it shows
a very pronounced habitual resemblance. The position of the
leaves are exactly the same, distichous with the blade held
edgewise ; besides the inflorescence is very rich-flowered. How-
ever, while we examined the anatomy we became aware of the
fact, that in this respect our plant showed a leaf-structure dif-
ferent from that of /. autumnalis, but almost exactly like
that of /. thermalis. In order to secure correct identification
of this South American species, we sent specimens to Mr. CO.
B. Clarke at Kew, who kindly replied that the plant was Link’s
LF. complanata var. Kraussiana. If Dr. Boeckeler had known
of the structural divergencies between this and /. autumnalis,

Pcl)
Geek

a.
Ya} 42)
Ld @&, oe
aa: a
==

SS
Re

BGS We Fie. 8. Fig. 9.
Leaves of F, ciliatifolia, F. Warei and F. sienophylla; transverse sections. x 75.

he would hardly have considered them as but one species,
even if the outer likeness is very striking. : |
If we now compare the leaves, of which we have drawn:
some transverse sections from the middle part of the blade, it
is readily seen that the outline and the structure of the upper
epidermis presents certain features for distinguishing the
species. Nevertheless these leaves demonstrate several inter-
gradations, thus the various forms may be easily deducted
from one another. The epidermis of the upper surface is
either developed as one single layer of large, bulliform cells,
or as several strata, that cover. the mesophyll. None of these
strata are, however, differentiated in such a way as to have the
function of opening or closing the leaf, the so-called “ Gelenk-
cellen” of the Germans, common to Cyperacew and Graminee.
In passing, to describe the minor details of the leaf-structure,
we observed the cuticle to be very thin and smooth in J
autumnalis, F. lava and FP. complanata, less so in F. spadicea ;
in the remaining species the cuticle is very thick, but smooth,
not granular. The epidermis of the lower surface is almost

T. Holm—Studies in the Cyperacee. 443

~ smooth with no furrows, and consists usually of smaller cells
than that of the upper, and is frequently developed into uni-
cellular hairs or very .minute spines, which are directed
upwards. The stomata are restricted to the lower epidermis,
where they form longitudinal rows outside the mesophyll
between the mestome-bundles. Characteristic of the stomata
is the deep and somewhat narrow air-chamber, which we have
noticed in most of our species with the exception of /”. cas-
tanea, where it is shallow and wide. The guard-cells are pro-
jecting in /. autumnalis, F. laxa, F. puberula, F. spadicea
and F. complanata, but not in the other species; they are
extremely thick-walled in /. castanea.

Similar to all'the other Cyperacew, examined so far, silicious
cones are observable in the epidermis-cells that cover the
stereome. As regards mechanical support most of our species
possess a well-developed stereome, the location of which can be
seen on the sections, figured above. It accompanies the larger
mestome-bundles as hypodermal groups on the lower surface of
the blade, while on the upper it is restricted to the margin. A
few thin-walled stereome-cells may be found above some of the
largest mestome-bundles, but separated from these by meso-
phyll. Although the stereome does not form any sheaths
around the mestome-bundles, it nevertheless renders a strong
‘support to the leaf by its very thick-walled cells, and is alto-
gether very well represented in our species of /imbristylis,
with the only exception of / autwmnalis and F. laxa, in
which it is relatively thin-walled.

The chlorophyll-bearing tissue, the mesophyll, occupies only
a small part of the leaf-blade, since it is generally traversed by
a number of mestome-bundles, which are especially large in
the narrow-leaved species, such as /. capillaris, F. Warez, ete.
It forms a homogeneous tissue on both faces of the leaf, and
the cells are true palisades, arranged radially around the
mestome-bundles. Furthermore this tissue is very compact in
our species excepting in /” laxa, of which some specimens
from Florida showed wide lacunes; and /’ castanea, in which
the palisade-cells are almost star-shaped, leaving numerous and
quite large intercellular spaces. Tannin was observed to be
very plentiful in this tissue, especially in / spadicea, I.
puberula and F. thermalis.

Considering the mestome-bundles, these are all arranged in
a single plane in the leaves of /imbristylis ; they vary in pro-
portion to size and outline, the larger being more or less oval,
the smaller constantly orbicular. They are imbedded in the
mesophyll, and are, as we have seen above, in no connection -
with the stereome, hence they are “ pure” mestome- bundles,

444 T. Holm—sStudies mm the Cyperacee.

Besides being surrounded by the radially arranged palisade-
cells, the mestome-bundles possess, also, a completely closed
parenchyma-sheath of small, thin-walled cells, which contain

chlorophyll. Inside of this parenchyma-sheath there is a

mestome-sheath, thin-walled in 4. autuwmnalis and J. thermalis,

but more or less thickened in the other species. The peculiar

“inner chlorophyll-bearing sheath,” which we. described in a
previous article upon Lzpocarpha,* occurs, also, in Hembris-

tylis, and not only in the genus “sensu strictiori,” but also in —

those American species of /solepzs which we have included in
the former genus. We found the structure of this sheath to
be identical with that of Lepocarpha, being composed of much
larger cells than the mestome-sheath; also by being closed in
the small, orbicular bundles, but interrupted in the larger
(fig. 10). The cell-content seems,
also, to present the same deep
green color as noticeable in Lepo-
carpha. 3
As regards the hadrome and lep-
tome these have, naturally, attained
their highest degree of develop-
ment in the largest mestome-
bundles, where these tissues are
sometimes separated from each
other by one or more layers of
thick-walled mestome-parenchyma,
as in [1 castanea, F. puberula, F.
spadicea and FL. complanata. The
presence of the inner, chlorophyll-

Limbristylis will necessarily place
the genus as a member of Rikli’s
“ Ohlorocyperacee,’ among which
Fig. 10. Transverse section of this author has not, however,
aan ae Be eee e enumerated any of the Jsolepis-
the upper surface. x 400. species examined by him. More-
over the morphological and ana-

tomical diagnosis of Zsolepis, as demonstrated by Palla, does
not seem comparable with any of the North American plants
which by Torrey were referred to this genus. Our species
differ very much from J/solepes proper by their morphological
characters, also by their internal structure. And even if the
inner, green sheath is not a constant character to all species in
certain large genera such as Heleocharis and Cyperus, as dem-
onstrated by Rikli, it would not be correct to consider our

* This Journal, vol. vii, p. 172.

bearing sheath in our species: of—

T. Holin—Studies in the Cyperacee. 445

American species as only anatomically distinct from other rep-
resentatives of /solepis. Our narrow-leaved species of /%m-
bristylis do not show any very important anatomical or
morphological characters by which they should be considered
as distinct fromm the others with somewhat broader leaves;
they do not present any such divergencies which induced Rikli
to separate his Chlorocyperus and Chlorocharis from Cyperus
and Heleocharis ; besides they lack the very conspicuous char-
acter: “‘ bractea infima erecta, inflorescentiam quasi lateralem
seepissime excedens, caulem continuans” as pointed out by
Bentham.

The stem.

The stem above-ground, which bears the inflorescence at its
apex, varies from cylindrical to being almost flattened, but
shows, nevertheless, a very uniform structure in our species.
The singularly flattened stem of /! autumnalis (fig. 11) and
F’. complanata corresponds well with the flat leaves, which are
turned edgewise in these species. In / castanea, F- thermalis,

Fig. 11. Transverse section of the stem of F autumnalis. x60.

F.. spadicea and FL. puberula the outline is almost cylindrical ;
in J”, laxa it is polyedric; in /. Warei and FP. stenophylia the
outline is decangular, while / capillaris possesses a penta- or
hexa-gonal stem. When compared with the leaf the stem-
structure shows many analogies, but the mechanical tissue has
here attained a stronger development, the mestome-bundles are
arranged in several, more or less concentric, bands and the
bark-cells are not arranged radially around the mestome-bundles
in all the species.

While the cuticle of the stem shows the same slight varia-
tion as we noticed in the leaves, the epidermis is often more
thick-walled and does not develop such dense pubescence as in

i: i
;
ia

it) }
}

1 1K

> ne
4

- ry

oat

AY ttt

“|
at
Wi

i
|
|
bales

aes

446 T. Holm—Studies in the Cyperacee.

the leaves. Grooves or furrows are observable in some species,
but they are rather shallow and wide; thus the stomata, which
are located in these, are freely exposed. We notice the same
variation in the relative width of the epidermis-cells as in the
leaves; besides the characteristic cones are very numerous out-
side the hypodermal stereome. ‘The bark-parenchyma forms,
similar to the mesophyll, a homogeneous tissue of palisade-
cells, which are arranged radially towards the center of the
stem in the majority of our species, but not in /. capzllaris,
Lf. ciliatefolia, F. Warer and F. stenophylla. The bark-
parenchyma is commonly quite solid with no lacunes, but the
intercellular spaces are numerous and wide in / castanea, L.
thermalis and F. puberula. It is, furthermore, to be pointed
out that the bark-palisades, which border on the leptome-side
of the innermost mestome-bundles in /” castanea, do not
radiate towards the center of the stem, but towards the center
of the bundle. Cells containing tannin were, also, observed in
the stem of various species, but not so abundant as in the
leaves. We have stated above, that the stereome is more
strongly developed in the stem than in the leaves. This is
readily noticed if we compare the leaf-section of /. castanea,
for instance, with the accompanying figure of a part of its
stem (fig. 12). This section shows a number of very large

va

Ke)
(AS a “4
“0 t ie \) OES0
40% [? 5 Eee
Sey eal : F.

“Saitse

oF

va

&
sy

Fig. 12. Transverse section of half of the stem of F. castanea. x5.

hypodermal stereome-bundles, most of which are situated deep
into the bark between the outer band of mestome-bundles.
There are, also, a few scattered and very small groups of
stereome on the hadrome-side of the innermost bundles.
Finally in the pith itself we observed quite a number of
stereome-bundles, which were not arranged in any order, and

ay

Vs
, 2
é
“7
any
*
M
=

ote

4
q

T. Holm—Studies in the Cyperacee. 447

of which only a few surrounded minute leptomatic-bundles.
Similar stereome-bundles imbedded in the pith were, also,
observed in /. spadicea, but not in any of the other species,
where this tissue is either strictly hypodermal or may be found
on the hadrome-side of the innermost mestome-bundles. The

largest part of the stem is occupied bya solid pith, mostly con-

sisting of polyedric cells, when considered in transverse sec-
tions. With the exception of /. spadicea, of which the pith
contained much tannin, this matter was not observed in the
pith of any of the other species.

In regard to the mestome-bundles these are more numerous
in the stem than in the leaf; they may form one single and
almost concentric band as in /. capillaris (fig. 13), F. Wares,
fF, ciliatifolia and F”. stenophylla, or several as in the other
species. They show exactly the same variation as those in the
leaves, and the smallest, the orbicular ones,
are invariably the most numerous and located
near the periphery. Inside of these may be a
second band of somewhat larger bundles as in
F. autumnalis, F. puberula, F. complanata,
Ff laxa and F. spadicea, while even a third
band is present in /. castanea and F. therma-
lis, these two species being the most robust of verse section of the
this genus in North America. The minor stem of F capillaris.
structure of the mestome-bundles is the same, * ':
all possessing a parenchyma-, a mestome- and an inner, chloro-
phyll-bearing sheath ; besides the relative development of the
leptome and hadrome corresponds with what has been said
above in the study of the leaf. There are furthermore some
leptomatic bundles located in the pith, as mentioned above
(fig. 14), but these were only found in one species, /. castanea.
These small bundles were surrounded by a few layers of thin-
walled stereome, aJso by a sheath, that
evidently may be identical with a
mestome-sheath ; no inner chlorophyll-
bearing sheath was observed here, since
these bundles occupied an almost cen-
tral position in the pith.

The peduncle.

The anatomical structure of the
peduncles, which bear the spikes, does Fig. 14. A leptome-bundle,
not differ in any essential degree from Lakes coaemld drape test 38
that of the stem. In F- castanea, for yond sootign x 320,
instance, the peduncle is slightly com-
pressed ; the pith occupies a smaller part of the section and

has no bundles neither of pure stereome or with leptome.

eA sa

oO a
——

—
_

» * . ‘ ~
wo Sa : Ls
>.) — ras

_— - =—_ a = ==

_
~

M
7

"eS * * — - . > - o _ . ) = ’ r = ww .
~ ee : bee A ee —— ‘ —_ B > oe, ee = ak: ye . a ee : ee
— ; i La “am - ec _ - . - 3 ’ - » — -
- - 2 — Soe » ’ > .
: : = ==: = — eee, ca oe 2 a= E az =
=e = = = << = = a = r ——— rr os ——— —— > = a SSS a

5 oe

448 T. Holm—Studies in the Cyperacee.

Tannin was found in both the bark and pith. The samereduc —
tion in the size of the pith was, also, observed in /. ava and —
Ff’, autumnalis ; in the latter the stereome appeared to have —
become more developed than in the stem, forming larger

groups and, also, by being more thick- walled.

The root.

With the exception of Van Tieghem,* Russow, De Bary
and Klinge, very few authors have given account of the root- —
structure in the Cyperacew. The elaborate work of Klinge,
to which we have often referred in our articles upon this order,
contains a number of detailed descriptions, by which we are
enabled to obtain a broad view of the general structure besides
a comparison between the principal characteristics of the roots
in Cyperacee and Gramineew. The very important character
to which Klinge has called attention as one of the most reliable
for distinguishing roots of Cyperacee from Graminew, the
tangential collapsing of the bark-cells, this character has, also,
been observed by the writer ina number of genera from this
country, including Fimbristylis. The structure of the endo-
dermis seems, according to our own observations, to represent
a number of variations. ‘This sheath occurs in our species of
Fimbristylis with more or less thickened cell-walls; in / cas-
tanea the thickening appears to have reached the maximum,
and the individual cells look as if their lumen had almost
entirely disappeared. In /. thermalis and F. stenophylla the
endodermis is, also, distinctly thick-walled with a number of
layers repr esenting an O-endodermis. And in /. autumnalis, —
which shows a rather weak mechanical structure, the endo-
dermis is, nevertheless, somewhat thickened, sufficiently to be
characterized as an O-endodermis. In F’. lawa the endodermis-
cells are not thickened all round, but only on the radial and
inner walls in the shape of a typical U-endodermis.

In regard to the pericambium, we have only noticed this
tissue as consisting of a single layer of thin-walled cells, or as
in F. castanea of thick-walled. Otherwise the pericambium
shows the same peculiarity, discussed in the works cited above,
by Klinge and Van Tieghem: that it does not form a closed
ring in all the Graminew and Cyperacee, but is frequently
interrupted by the protohadrome. It will, however, require
further observation in order to ascertain, whether this fact is
common to all or only to certain genera of the Cyperacee.
While Klinge has observed the protohadrome to border
immediately on the endodermis, thus breaking through the

* Van Tieghem, Ph., Recherches sur la symétrie de structure des plantes vascu-
laires (Annales d. sc. nat., 5th series, vol. xiii, Paris, 1870-71, p 5).

TL. Holm—Studies nm the Cyperacee. 449

pericambium, in several species of Carex, Hriopharum, Scirpus,
feleocharis and according to Duval-Jouve, also in Cyperus
globosus, C. fuscus, C. vegetus and C. serotinus, Van Tieghem
did not observe this to take place in Carex brizoides. More-
over both Duval-Jouve and Klinge observed a similar separa-
tion of the protohadrome from the endodermis in various
species of Cyperus and Galilea mucronata. In Carex Fraseri,
as we have mentioned in a previous article, the protohadrome
is, also, separated from the endodermis. The innermost part
of the central-cylinder is in /. autumnalis, FE. laxa and F-.
thermalis occupied by a single, large vessel, and several in /.
castanea and FL. stenophylla. The conjunctive tissue seems
most often to be thin-walled, with the exception of /” castanea
and F. Jaxa. ‘The structure of the roots in our species of the
genus does not seem to present any other mechanical protec-
tion than that rendered by the thick-walled endodermis,
although we should have expected to find a similar protection
in the innermost layers of the bark corresponding to what we
observed in Scleria and Lipocarpha with their thick-walled
bark-parenchyma.

These anatomical details, which we have observed in North
American species of /vmbristylis, seem to illustrate that we
have only one genus before us; the species which by Torrey
were referred to /solepis appear to be inseparable from the
others, not only from an anatomical, but, also, from a morpho-
logical viewpoint. And if we consider the various species, as
we have studied them in this country, they exhibit a sufficient
number of characters which may be regarded as “ mutual
affinities,” such as these are understood in orders as large as the
Cyperacee. But it does not seem possible to draw any such
distinction, as for instance to separate the species into “ Xero-
phytes or Hydrophytes,” even if the local environment might
suggest these plants to belong to different “ plant-societies.”
Because the internal structure of the stem, the leaf and the
root does not furnish any evidence that certain anatomical
features are necessarily dependent on a special climate or soil,
neither very dry nor very damp. It seems much more natural
simply to acknowledge that such and such species live under
such and such conditions, even if we, on the other hand, are
unable to account for the similarities and diversities in struc-
ture. The statement that bog-plants may possess xerophytic
characters and vice versa gives no satisfaction, and is even mis-
leading; characters may have been simply inherited without
being at present of any particular advantage to the species.
The inner, chlorophyll-bearing sheath may represent a char-
acter of that kind, its function being, so far, unexplainable.

450 Ts Holm—Studies in the Cyperacec.

This sheath is, as we have stated above, not especially charac-
teristic of any type of /imbristylis; it is common to our
native species and a number from the old world. It is not
restricted either to any certain shape of leaf or to species from

any peculiar region or locality. Furthermore the relative

development of the mechanical tissue in our genus from dry
as well as from wet soil furnishes no information regarding a
supposed existence of plant-societies with xerophytic or hydro-
phytic structures ; neither the radial arrangement of the pali-

sade-cells or the presence of a parenchyma-sheath. Hence an

attempt to classify our native species of %mbristylis as

“xerophytes or hydrophytes” would seem very unnatural

and not in conformity with their biological peculiarities.
Brookland, D. C., December, 1898.

5
3
2
F,
-
.
{

a es, OY ee ee
adh iy ;

-_ te De

Hillebrand, Turner and Clarke—Roscoelite. 451

Art. XLIX.—On Roscoelite; by W. F. HILtLeBranp and
H. W. Turner, with a note on its chemical constitution by
F. W. CLARKE.

ANALYSIS AND Composition, By W. F. HILLEepRanp.

THE rare mineral roscoelite has greatly needed reéxamina-
tion in order to reconcile the discrepancies between the analyses
of Roscoe and Genth and to establish a satisfactory formula
for this supposed vanadium mica. To the kindness of Mr.
G. W. Kimble of Placerville, Cal., I am indebted through
Mr. H. W. Turner for specimens from the Stockslager mine,
from which a limited amount of fairly pure material was
picked out. This was then laboriously purified by the aid of
Thoulet’s solution, the result being a very nearly pure product
weighing only 1:2 grams and having after drying at 100° C. a
density of 2-97 at 30° ©,

Notwithstanding the small amount, it was possible by the
exercise of care to make fairly satisfactor y analyses.

With regard to the methods employed little need be said
except as to the determination of the condition of the vana-
dium. For this purpose decomposition was effected by rather
dilute H,SO, in sealed tubes, the greatest care being taken to
expel every trace of air from the powder and acid and to seal
the tube during passage of a current of CO,. Otherwise it is
impossible to prevent oxidation of a considerable part of the
V,O,. In one ease, unfortunately, the air cannot have been
fully expelled, for the solution after decomposition was blue
instead of green and much less oxygen was required in titra-
tion than when the color was green.

The contents of the tube, still warm, were poured into
fairly hot freshly boiled water and titrated rapidly. Iron and
vanadium were then reduced by H,S gas, the latter boiled out
in a current of CO,, and titration repeated on the hot liquid.
The solution was then boiled with ammonia, the precipitate
fused with Na,CO,, leached with water, and the residue again
fused with Na,CO, and leached to remove the last of the
vanadium. This residue was then fused with KHSO,,* dis-
solved in dilute H,SO,, boiled first with H,S and then i in a cur
rent of CO,, and the liquid titrated for total iron. The solu-
tion held titanium which was then estimated colorimetrically.

The first of the titration results gave the effect of all iron,
assuming its existence as FeO, and of all vanadium that might

* Any slight trace of vanadium remaining will impart a bright yellow color to

the cold KHSO, fusion, a test which proved useful more than once during the
analysis.

Am. Jour. Scr.—Fourts Series, Vou. VII, No. 42.—June, 1899.
30

452 Hitlebrand, Turner and Clarke—Roscoelite.

exist in a lower state of oxidation than V,O,. The second
gave all iron as FeO and all vanadium as V,O,. Deduct from
both the figure for FeO and the remainder gives that for
vanadium. In this way two very concordant results were
obtained for total vanadium as V,O, which were supplemented
by tests on portions used for other constituents, but only one
was obtained for the vanadium as it exists in the mineral, a
second being vitiated by evident oxidation during decomposi-
tion in the tube. As a check, however, a fresh sample of
unpurified mineral was similarly treated and it was found that
fully nine-tenths of the vanadium existed as V,O,, a result
confirming the single test on the purified material whieh
showed 93°5 per cent as V,O,. It is not impossible that slight
oxidation had taken place even in these cases, and I feel justi-
fied in assuming with Genth that the vanadium should be con-
sidered wholly as V,O,. |

In the other portions analyzed the vanadium was likewise
titrated in V,O, condition, but only after separation from iron,
titanium and aluminum by fusion with Na,CO,, extraction
with water and separation of dissolved alumina by ammonium
carbonate. A second fusion of the residue and of the precipi-
tated alumina was necessary in order to extract all the vana-
dium. These numerous manipulations render the figures for
Al,O, perhaps the least trustworthy of all, but the average
given is probably not far from correct.

The iron is assumed to be present as FeO; and the titanium
to belong to a foreign mineral, since a test on unpurified mate-
rial gave much more, namely, 1°50 per cent TiO,, without
accompanying increase in FeO, which latter observation seems
to exclude ilmenite as the source of the titanium.

Both the iron and magnesium are supposed to belong to the
roscoelite, since they were found by Genth in nearly the same
amounts and no recognizable iron or magnesium minerals were
noticed in the purified powder. | |

For comparison, the mean of Roscoe’s analyses and that one
of Genth’s considered by himself to be his best are also given’
in the table on the following page.

Very marked differences are apparent in the three analyses
by different chemists. If titanium was present in the mate-
rial analyzed by Genth and Roscoe, as is very probable, their
high results for alumina are in great part at least accounted for.

It is inconceivable how. Genth obtained his value for water by
ignition, since the mineral oxidizes when heated in air. In
fact the oxidation in one of my own analyses, after allowing ~
for loss of water as ascertained by direct weight, was almost
what theory requires for the oxidation of V,O, to V,O, and of
FeO to Fe,O,, or 5:14 per cent instead of 5-27 percent. It

= o>

ee NA ee i =
== = Ss = = =. = = =

—— Ses ==

= -
«=

7

;

:
a
4

5 oo. *
— - |
a2 SS A

= =

— eae a

y
ff
.
oh) Aly
A ie \
a* } ma
1 hd aa
* r t M
i J ot hit
:
r : -
| ol |
j
, : :

a Ee

\ -S 4

Hillebrand, Turner and Clarke—Roscoelite. 453

may fairly be assumed that his water was weighed directly
after expulsion by ignition of the powder. Roscoe’s figures

j for water, if not for moisture, must be affected by error, prob-
| ably arising from the unsuspected oxidation of vanadium.
f ANALYSES OF ROSCOELITE.
j | Hillebrand. |
Amount —— -~— ——— Genth. Roscoe.
used. | -3000g.| *2531g. |-2635g.|"1560g.*|-2038g.| Mean.
SS | 45°30 | 45-04 | 4517! 47°69 |Si0.._--] 41°25
a ‘ate k, cece 80] .°78 | VsOy Hib Sas 60t
Wee eesS. | 23°90 | 24:00 | 24:09 | 24:06) 24°01) 20°56/Al,0, __| 14:14
7.) Poe 11-74 | 11-34 | 1154| 14:10|Fe,0, __|. 1:13 -
(LE nanan 1°59 | | 1:65 157} 1-60 1-67 |Mn.G, |" 1:15
MgO _....| 1:64 | 164) 2°00 |CaO ____ ‘61
ee! | 110-41 = 10°32 | 10-37| 7°59|MgO .._| 2701
Ee) ! | 13 | none | b=!" O6 ci A ahd 2s, geen 8°56
fe) eee ) it. tr / _ ft. tr. ie... |NasOr >= 82
H.0O below | Water __ 1:08
iia Us} (+ -40 2 34.6 Moisture rH |
H.O 105-. | | z
280°C. _| 17 en 17) 101-62
H.0 above | \ | 4°96 |
280° C._.| 4:12) ign.
1 | ae aoe / _ none |
| s |
| | | 99°86 98-76)

From the column of means of my own analyses the follow-
ing ratios are obtainable :

lo ae ee eels hag 753
gs aa SSI 159
Pet Os eee 113
MeO 3 5 eee 022
MeQ: 2.4 ier eet 041
MeO) tc eee
BO eo. cue 229

The entire absence of manganese and of calcium in my own
and Genth’s samples tends to confirm the suspicion that Ros-
coe’s material was far from pure. It is to be remarked, how-
ever, that my figures for vanadium agree quite closely with his
and differ widely from Genth’s.t

Discrepancies of this kind are not necessarily to be ascribed
to faulty analyses. {[t is well enough known that in any one

* 4-94 per cent oxygen used for complete oxidation instead of 5°27 needed
for all V as V2.0; and Fe as FeO:

+ Equivalent to 23°59 per cent V20s.

¢ This agreement is probably more apparent than real for the reason that
Roscoe’s figure would be materially lowered by regarding the iron of his analysis
as FeO instead of Fe.O; in conformity with my analysis.

454 Fillebrand, Turner and Clarke—foscoelite.

species of mica various molecules must sometimes be assumed
to exist in different proportions and the general formula for
such a species can only be arrived at by comparison of a series
of analyses of different varieties. Hence, in view of the lack
of any simple ratios, the deduction of a definite and final
formula from my data is not justifiable. Further analyses are
needed of new and very pure material from other locations
even if these be not far removed from the source of the
present material. Nevertheless, in the hands of an expert very
unpromising data may often be made to afford positive indica-
tions, and that this is true in the present case the following dis-
cussion by Professor Ff. W. Clarke clearly shows.

CHEMICAL CONSTITUTION OF RoscoELiTtE, By F. W. CLARKE.

The ratios given in the foregoing new analysis, used directly,
lead to the following empirical formula for roscoelite :
HK Kees Al Veo 0

458 220 153 ~ 2724°

Here H to K, and Mg to Fe are as 2 to-1. Between O and 8i,
however, the ratio is not simple, and lies below the orthosili-
cate and above the trisilicate proportion. Since in many micas
the groups SiO, and $i,O, are replaceable, that suggestion may —
be followed out here? and then the formula reduces to

R’ aR" A) by pelCoike We es 1, One

From this expression, applying Clarke’s mica theory, the min-
eral may be regarded as a molecular mixture of the three

compounds
He 2. 3.
yp bl0= Melt 21,0 we a0, = ee
Al—Si0O,= MgH Al—Si,0, = KH, Al—S8i0, = V
™S10, = MeH, ™Si,0,=Al, . ™810,= V,

in the ratio 21: 22: 159, or nearly 1:1:8. Upon reducing the ~
analysis to 100 per cent, after throwing out the TiO, and the
water lost below 280° as extraneous, we get the following com-
parison between the results found and the theoretical compo-
sition.

Found. Reduced, Calculated.

S10, ee aa 45°17 45°88 45°52
TiO; ethos ee 78 peck abe
V,0, ae alias “Y iapueeee os 24°01 24°39 24°64
Al,O, OS ee 11°54 a ares} 11°62
MgO chet h 3 58 dba aes 1°64 1°66 re
BeO pee ee oneleGo 1°63 1°35
K,O aa Seal ea 10:37 10°53 10°81
EO; O80 yi oy, Sa eee

eo ADB O uelween Oe ae 4°18 4°14

99°80 100°00 100°00

z
bs
a

| Hillebrand, Turner and Clarke—foscoelite. 455

This comparison, based on the ratio 21 : 22: 159, is as satisfac-
tory as could be expected.

Of these component molecules, the first represents the normal
phlogopite type: the second is a trisilicate alkaline biotite;
and the third, which forms 74°5 per cent of the whole mass, is
a muscovite in which two-thirds of the aluminum have been
replaced by vanadium ; in short, a vanadium muscovite. Ordi-
nary muscovite is Al. (Si0,), KH, ; and whether a correspond-
ing V,(Si0,),KH, exists, can be “determined only. by analyses
of roscoelite from other localities, and so learning its range of
variation. That vanadium may replace aluminum is shown by
the fact that Piccini has prepared true vanadium alums. That
roscoelite is essentially a vanadium muscovite seems to be fairly
well established. As for the molecule AJ,($i,O,),K,H,, its
existence is indicated in some other micas; and in Simmler’s
“helvetan ” it seems to be the dominant molecule.

THE OccuRRENCE oF RoscoELirE, By H. W. TuRNER.

According to H. G. Hanks,* at one time State mineralogist
of California, attention was first called to roscoelite by Dr.
James Blake at a meeting of the San Francisco Microscopical
Society, July 2d, 1874. The specimens then exhibited were
from the Stockslagert mine, which is about 1" southwesterly
from Lotus on Granite Creek in Eldorado County.

At a meeting of the California Academy of Sciences held
July 20, 1874, Dr. Blake presented specimens of the same
mineral, which he then supposed to be a chromium mica.

At a meeting of the California Academy held August 2,
1875,{ Dr. Blake read a paper on roscoelite. Samples sent by
him to Dr. Genth of Philadelphia were found to contain
vanadium. ‘The mineral was named by Blake in honor of
Professor Roscoe of Manchester, England, who had made
vanadium a special study. Dr. Blake calls attention to the
fact that Dr. Hall found vanadium widely diffused in many
rocks, generally associated with phosphorus. According to
Hanks the Stockslager vein from which the roscoelite obtained
by Dr. Blake was taken, is small and not continuous, varying
from two inches to a foot in thickness and running nearly
parallel with Granite Creek. Associated with the quartz is
calcite, and there are at least two varieties of iron sulphide
present, probably pyrite and chalcopyrite. Hanks states
further that the gold occurs only with the roscoelite and is

* Second Ann. Rep. State Mineralogist of California, pp. 263-4, 1880-2.

+ According to Lindgren (see Economic Sheet of the Placerville folio of the
Geol. Atlas of the U.S.) this name should be spelled as above, but Hanks spells it
Stuckslager.

+ Proceedings Cal. Acad., vol. vi, p. 150, 1875

456 Eillebrand, Turner and Clarke—foscoelite.

found interstratified with the roscoelite lamine in pieces from

the value of one dollar to the minutest microscopic particles.

He also states that from four to five hundred pounds of ros-
coelite were obtained by the miners, all of which was wasted in
extracting the gold.

Mr. George W. Kimble of Placerville, California, for many
years county surveyor, has furnished the California material
analyzed by Dr. Hillebrand as well as other specimens. In
these specimens the roscoelite is in part embedded in quartz
and probably contemporaneous in formation with the quartz,
and in part fills little cracks in the quartz and therefore some-
what later.

There are given below four localities where roscoelite has
been found, according to Mr. Kimble, to whom the author is
indebted for the following information about the occurrence of
the mineral at these localities.

Ca\averas Sormation, Ser pent: ne
Bene rite but isre.

Fie. 1. Gentanien map copied from the Placerville folio, showing roscoelite
localities. Geology by W. Lindgren.

The accompanying map, on the scale of about 1-150000, gives
the exact localities where the roscoelite has been found. The
Calaveras formation is largely of Carboniferous age, as may

Hillebrand, Turner and Clarke—Roscoelite. 457

also be the augite-porphyrite tuff associated with it. The ser-
pentine is later than the Calaveras formation and earlier than
the granodiorite, which is probably late Jurassic or early Cre-
taceous. It will be observed that all of the localities are at or
near the contact of an intrusive granodiorite mass with the sur-
rounding rocks, chiefly sediments and older lavas. This suggests
that the mineral may be regarded as in some way due to the
mineralizing solutions accompanying the intrusion of the gran-
odiorite. However, the quartz veins clearly fill fractures
which formed in the granodiorite and associated rocks after the
consolidation of the granodiorite ; consequently the deposition
of the quartz and the associated gold and roscoelite must have
been also subsequent to the consolidation of the granodiorite.
The following are the localities reported by Mr. Kimble.
They are all in Eldorado County, from 8 to 15*'™ northwest of
Placerville.

Locality 1. Thompson Hill on itsnortheast slope about 2:55"
southeast of the Stockslager mine.— There are here fif-

teen or more small seams of quartz having a strike of north

of west. As these seams pass through the contact of the
granodiorite with the greenstone (augite-porphyrite?) they con-
tain rich spots of gold and roscoelite. The seams of quartz
pass on through the northeast point of Thompson Hill and
come into the granodiorite again.

Locality 2. WStockslager mine on Granite Creek. — Mr.
Lindgren informs me that the vein of the Stockslager mine is
in a narrow wedge of metamorphic sediments of the Calaveras
formation. Immediately east of this wedge is granodiorite,
and immediately west is serpentine. Mr. Kimble states that
there is here but one vein of quartz. This passes from the
granodiorite into the narrow wedge of the Calaveras formation.
It does not penetrate the serpentine.. As previously stated, it
was at this locality that by far the larger part of the roscoelite
was obtained, including that analyzed by Dr. Hillebrand,

Locality 3. South slope of spur about 2-78 north of the
village of Lotus.—In micaceous slate, which is part of a con-
tact metamorphic zone of the Calaveras formation. The exact
locality is about 500 feet east of the contact. It has never
been demonstrated, but Mr. Kimble thinks that this quartz vein
extends southeast into the granodiorite.

Locality 4. Clark Mountain on its east slope at the contact

of the Calaveras formation and the granodiorite.—The oceur-

rence here is identical with that at the Stockslager mine.
According to Kimble no roscoelite has ever been found
where the seams are in granodiorite, and the latter is not
altered; but some has been found at the contact of the quartz
with serpentine in localities 2and4. All of the quartz seams
at localities 1, 2, and4 dip southwest and have pockets of gold.

7 <2 a < - ad . a = —<—_ =
~ > eee - 3 bh F = : ~ + ~ omer rr s a " —
= = ar w er ——= ’ “= Set LS eo =. a b =. a
aia = er < = Se ie pe, a 2ST = —— = = ——— — —
a i — 3 == - = S = — = —— = e - =

if

458 Hillebrand, Turner and Clarke—Roscoelite.

Prof. Hanks* states that roscoelite was also found in See. 31,
T. 11 N. Range 10 E, two miles from the Stockslager mine.
This appears to be the Thompson Hill locality of Kimble, to
whom, indeed, Prof. Hanks was probably indebted for his
information. Hanks states that the roscoelite was found here
in the bed rock of Big Red ravine in a dark-colored micaceous
rock in small seams of quartz with calcite and gold.

The roscoelite from California shows a tendency to ecrystal-
lize in little rosettes, so that mdividual scales of any size with
the same optical orientation throughout are difficult to obtain.
Some scales gave with convergent light a nearly uniaxial black
cross, the hyperbolas opening but slightly on rotation, indicat-
ing a small axial angle. Like all other micas it is optically
negative. The pleochroism as seen in thin foils is, ¢ and 6 clove
brown to greenish yellow brown, a light greenish yellow. In
the scales of the powder analyzed no rutile needles or other
inclusions were detected, but in the center of a rosette of ros-
coelite scales were found dark grains with a metallic luster.
These grains are of sufficient size to have fallen with the heavy
minerals during the process of separation with the Thoulet
solution, and it is not likely that any of them have been

included in the powder analyzed, although they were noted in

the impure material before it had undergone the final separa-
tion. The grains show no erystal form and were not deter-
mined.

* Loco citato, p. 263.

Nipher—Gravitation in Gaseous Nebule. 459

Art. L.—On Gravitation in Gaseous Nebule; by FRANCIS
EK. NIPHER.

THis subject has received attention of late through the
work of Dr. See, who has rediscovered the law announced by
Ritter in 1878. According to Ritter, if R be the radius of a
spherical mass of gas of cosmieal dimensions, and T its tem-
perature, the product TR =constant. As Ritter announced,
the heat capacity of such a gravitating mass is negative. If
heat leaves the gas, it contracts and becomes warmer. Dr. C.
M. Woodward has recently published a paper* in which he
deals with the conditions of equilibrium in such a mass, and he
deduced the equation for the mass of the ¢entral core of a
gravitating spherical nebula

py i SAS
gis

Here C is the constant for the gas, and M is the mass
internal to any radius7. K is the gravitation constant and T,
is the constant temperature of the mass. Woodward admits
that if such a mass could be contracted, its temperature would
rise, but he denies that gravitation is competent to contract the
gas, and even concludes that such a nebula cannot lose heat by
radiation.

The present writer, in the succeeding number of the Trans-
actions, has shown that such contraction is possible; and that
when T is made variable in the above formula, the value of
Ritter’s constant is determined by that equation, in terms of
the gravitation constant, the constant for the gas, and the mass
M internal to 7.

This equation is then applied to a cosmical mass of hydro-
gen. What must be the physical condition in order that a
central mass or core, having a radius equal to that of the sun,
should contain a mass equal to that of the sun. The condi-
tions of the problem determine 7, OC and M; and K_ being
known, the value of T turns out to be 20,000,000 degrees cen-
tigrade. The pressure at the surface of this sphere is com-
puted, and it is found to be 3°706x10" dynes per square
centimeter, or 366,000,000 atmospheres. The average densit
of the spherical mass, which is three times the density at the
surface of the hydrogen sun, is about 7 per cent less than the
average density of the sun itself. The pressure at a distance
of 92 million miles from the center of mass is found to be
about 0:4 of an atmosphere.

M

* Trans. Acad. of Sci. of St. Louis, vol. ix, No. 3.

460 Nipher—Gravitation in Gaseous Nebule.

The equations show that the same pressures and densities
would hold for any other perfect gas. In fact, the above equa-
tion shows that if » and M are fixed, the product TC must be
constant for all gases. This carries with it the conclusion
shown by the other equations, that the average density of such
a solar mass is independent of the nature of the gas. This is
a matter of great significance, when taken in connection with
the fact that the real density of the sun is only slightly greater
than the density of the hypothetical hydrogen sun.

The real condition around our sun is, that increasing opacity
to radiation as one goes to levels of smaller radius, has retained
the heat within the dense nucleus. The rarefied external parts
of the solar nebula have parted with their heat and the tem-
perature throughout the mass has ceased to be uniform. Inter-
planetary pressures have been abolished. Hydrogen would
solidify at a distance of 92 million miles from our sun if away
from any large mass of matter. And this obliteration of cos-
mical pressure has almost wholly compensated the fall in tem-
perature of the sun from 20 millions at least to perhaps
10,000 degrees. |

The fact that a gaseous mass can apparently contract itself,
and heat up in some such way as it would do if it were com-
pressed by the action of some external system, is obviously of
profound significance. The clue that it seems to give concern-
ing the nature of gravitation is well worthy of the most serious
attention. Is not gravitation the action upon matter of a
system whelly external to matter ?

Gooch and Peters—Titration of Oxalic Acid. 461

Art. Ll.—The Titration of Oxalic Acid by Potassium
Permanganatein presence of Hydrochloric Acid; by F. A.
Goocn and C. A. PETERS.

[Contributions from the Kent Chemical Laboratory of Yale University—LX XXII. |

LOWENTHAL and LENSSEN* were the first to show that the
titration of a ferrous salt by potassium permanganate in the
presence of hydrochloric acid, according to the process of
Marguerittet is vitiated by the evolution of chlorine outside
the main reaction, and to point out that a remedy for the diffi-
culty is to be found in the titration of the ferrous salt in
divided portions, other equal volumes of the ferrous solution
being added to the liquid in which the first titration is accom-
plished until the amount of iron indicated by successive titra-
tions becomes constant.

Kessler{t showed the restraining influence of certain sul-
phates, of manganous sulphate in particular, upon the irregular
and undesirable interaction of the permanganate and hydro-
chloric acid, and Zimmermann,$ in apparent ignorance of
Kessler’s forgotten proposal, advocated the introduction of a
manganous salt, best the sulphate, into the ferrous salt to be
determined, thus accomplishing the purpose of the empirical
procedure of Lowenthal and Lenssen.

The tendency toward evolution of chlorine in the oxidation
of a ferrous salt by permanganate, as compared with the
absence of such tendency in the similar oxidation of oxalic
acid, in presence of hydrochloric acid, was explained .by Zim-
mermann on the hypothesis that an oxide of iron higher than
ferric oxide is formed as an intermediate product, and that
this unstable oxide is sufficiently active to break up hydro-
chloric acid as well as to oxidize more of the ferrous salt.
Quite recently, Wagner] finds explanation of the sensitiveness
of the hydrochloric acid solution of the ferrous salt in the
probable formation of chlor-ferrous acid (analogous to chlor-
platinic and chlor-auric acids), which suffers oxidation more
readily than hydrochloric acid under the action of the perman-
ganate. The protective influence of the manganous salt turns
apparently, as Zimmermann suggested, upon the initiation of
Guyard’s reaction, according to which the permanganate and
manganous salt interact to form a higher oxide of manganese
of a constitution approaching the dioxide more or less closely—
this oxide being capable of oxidizing the ferrous salt, but slow

* Zeit. f. Anal. Chem., i, 329. + Ann, d. Chim. et d. Phys. [3], xviii, 244.
¢ Ann. d Phys. u. Chem., exviii, 48; exix, 225-226.

§ Ann. d. Chem., cexiii, 302.

|| Maasanalytische Studien, Habilitationsschrift, Leipzic, 1898.

es ~
———- 3 4 -
a eS =. pany
== er: +

a sor ses Se

a

462 Gooch and Peters—Titration of Owalice Acid.

>S

to act upon the hydrochloric acid, or the chlor-ferrous acid of
Wagner. According to Volhard* the reaction of Guyard is
favored and hastened by heat and concentration of the solu-
tion, while it is delayed by acidity and dilution; but even in
solutions containing very little manganous salt and a consider-
able quantity of free acid the faint rose color developed by
the careful addition of permanganate ultimately vanishes until
every trace of the manganous salt is precipitated. Whena

considerable amount of the salt is present interaction follows —

immediately the introduction of the permanganate. Zimmer-
mann advocates the use of 4 grams of manganous sulphate
uniformly in titrations of a ferrous salt by permanganate, a
procedure to which Wagner gives acquiescence, though point-
ing out that a ninth of that amount is all that he finds to be
necessary. The excess of the manganous salt can do no harm
so long as the higher oxide, the product of interaction of the
manganous salt and the permanganate, is immediately reduced
by even traces of a ferrous salt, and this appears to be the case
at least within the limits proposed by Zimmermann and Wag-
ner. Thus we find, as shown in results of the accompany-
ing table, that so much as five grams of the sulphate may be
present in 135°™ of the liquid, containing about 5™ of hydro-
chlorie acid of full strength, without interfering with the regu-
larity of the titration; and the effect is trivial even when the
amount of manganous sulphate reaches ten grams. We find
also practical regularity of working when manganous chloride
is substituted for the sulphate, and in this respect our results
accord with those of Zimmermann and differ from those of

Wagner:t :

Total yolume

at beginning HCl KMn0,

of titration. Sp.gr.1:09. FeCle. =, N. MnSO,.5H.20. MnCl..4H.0.
em?. cm®, cm?, cem?, grams. grams,
135 10 25 21°70 If tee
om N37) 10 25 21°70 3° pape 6
135 10 25 21°70 5° PRE?
135 10 25 91s 7s SS iSE
135 10 23 GIGS 10° rae hi
145 20 25 2175 10° hes
175 50 25 21275 iO cet
135 10 25 21°70 eae he
135 10 25 21°70 a See 2
145 20 25 94°70 ee 2°
155 30 25 21 hp eee 3
165 40 25 91°70 Seoete 4°

In all cases, however, in which the larger amounts of manga-
nous salt are present, the end reaction is marked by the advent

* Ann. d. Chem, excviii, 318, 1879. t Loc. cit., p. 104.

ee ee 2 ee

4 Poe

ee a ne a eo

bie © atts de ei

EE eee ee ee

4 sh

®

8 OAC SY yr i. Fn oe eee

:

Gooch and Peters—Titration of Oxalic Acid. 463

of a brownish-red precipitate rather than the clear pink of the
soluble permanganate, and it is obvious that in case the solu-
tions to be oxidized were not active enough to act with rapidity
upon the product of the Guyard reaction, difficulty might fol-
low the failure to adjust the conditions more particularly.

It has been stated by Fleischer* and Zimmermannt+ that
hydrochloric acid interferes in no way with the titration of
oxalic acid by potassium permanganate. This statement, how-
ever, is not in accord with our experience; for we find that in
such titrations there is a small though real waste of permanga-
nate proportionate to the amount of hydrochloric acid present.
This. fact is brought out clearly in the comparison of experi-
ments of section A in the following table, in which no hydro-
chloric acid was present, with experiments B, in which
hydrochloric acid was present.

Temperature at beginning about 80° C.

Approximate Ammonium Variation from
volume at H.2SO; HCl oxalate mean of A taken
beginning “a> t Sp. gr. 1°09. ia N. KMn0,. as standard.

of titration. cm’, em?. em?. cm? em?.

A
200 i) beat 50 47°50 0°00
200 i) eles 50 47°50 0°00
200 10 hyn sie 30 47°50 0°00
200 10 Le ted 50 47°50 0°00
200 25 oe 50 47°50 0°00
200 25 ee 50 47°50 0°00
oa 10 2°5 25 23°80 + 0°05
4 150 10 2°5 25 23°90 +0°L5
; 150 10 a-Os 25 23°90 +0°15
| 150 10 10°0 25 24°00 + 0°25
500 +s) i Roe 25 23°80 +0°05
500 10 10°0 -25 24°00 + 0°25
500 10 10°0 25 24°10 +0°35

From these results it is evident that, though the error intro-

- duced by the presence of the hydrochloric acid during the

action of the permanganate upon the oxalic acid is small, it is
plainly appreciable. ‘The questions arise, therefore, first, as to
whether the secondary action of the permanganate upon the
hydrochloric acid may be prevented by the presence of a suit-
able amount of a manganous salt, and, secondly, as to whether
in this event the reducing agent,—the oxalic acid—is sufii-
ciently active, like the ferrous salt, to prevent the premature
establishment of an end color due to the Guyard reaction. The
latter question must naturally be settled before the former can

* Volumetric Analysis; Trans. by Muir, p. 71. + Loe. cit.

464 Gooch and Peters—Titration of Oxalie Acid.

be taken up. In the accompanying table are recorded the
effects of varying amounts of manganous salt in presence of
different amounts of sulphuric acid in the reaction of perman-
ganate upon oxalic acid. .

Temperature at beginning about 80° C.

Volume H.SO, Ammonium Variation
at beginning. Waa oxalate. MnS0O,.5H.,0. KMn0O,. - from

cm’. em?. zon em?®. grams. em?. standard.

( 130 5 25 dens she We Lo 0°00
130 9) a 0°0008 23°75 0:00

| 130 on 25 0°00382 230 0°00
130 5 25 ‘0°0160 ISTS 0:00
{ 130 5 25 I 23°70 —0°05
| 130 3) 25 2° 23°75 - 0°00
130 9) 25 25 23°60 —0°15

| 130 5) 25 3°0 23°40 —0°25

| 130 5 25 4°0 23°60. . —0°15

( 500 3) 2am eta 23°80 +0°05
500 5 25 0°0008 23°80 + 0°05

| 500 5 ao 0°00382 23°80 28220505

< 500 5) 25 ile 23°70. ost 00s

1 500 5 25 24S 23°40 —0°35

| 500 5 25 3" 23°50 —0°25

| 500 5 20 4° 23°30 — 0°45

( 180 ( 10 25 ie 23°80 + 0°05
130 4 10 25 Day Paes we 0°00
130 | 10 25 ea, 23°65 —0°10
es 10) | 10 25 4° 23°50 — 0°25
30 15 25 Oe 93°15 0°00

| 130 15 25 4° 23°70 —0°05
130 15 25 5° 23°50 —9°25
130 30 25 ae PS tT a 0:00
130 30 25 4° 23°70 —0'05
(130 30 25 5B 23°75 0-00

From the results given it is evident that the persistence of the _

Guyard reaction is liable to interfere with the end reaction of
oxidation of oxalic acid unless an adjustment is made between
the quantity of the manganous salt, the amount of acid, and
the dilution. In hot solutions of a total volume of 130°%™* at
the beginning, no more than 2 grams of the manganous sul-
phate should accompany 5 to 10°™* of the 1:1 sulphuric acid;
when the total volume at the beginning reaches 500™*, no more
than a single gram of the salt should be present with 5°™ of the
1:1 sulphuric acid. The amount of manganous salt may,
however, be increased considerably if the quantity of acid is
increased.

7

Gooch and Peters—Titration of Oxalic Acid. 465

As Kessler has noted, a sufficiency of the manganous salt,
acting no doubt as the medium of transfer of oxygen, may
bring about interaction between the permanganate and the
oxalic acid at atmospheric temperatures without the tedious
delay ordinarily encountered in the attempt to consummate
that action in cold solutions. It would seem natural that the
manganic hydroxide formed in the Guyard reaction at low
temperatures should yield more readily to the reducing action
of the oxalic acid than the more anhydrous form to be expected
in hot solutions, so that at such temperatures the limits as to
proportions of manganous salt, acid, and dilution, within
which favorable action may take ‘place, ‘should be wider ; ; more-
over, the undesirable action of the permanganate upon “hydro-
chloric acid, when that acid is present, should be less appreci-
able at lower temperatures. In our experiments, therefore,
upon the oxidation of oxalic acid by potassium permanganate
in presence of hydrochloric acid, we have studied the effect of
varying the proportions of the manganous salt both at atmos-
pheric temperatures and the higher temperatures generally -
employed. ;

Temperature 20°-26° C,

| Ammo-

Number | Volume HCl nium Variation
of at begin- H.SO, Sp. gr.| oxalate MnS0O,. MnCl..| from
experi- | nipngof |/ 1:1 | 1°09. | #A,N. | KMnOQ,.| 5H.O. | 4H.20. |standard.

ment. titration.) cm?. | cm’*. en. em’, | grams. |grams.| cm?.
1 (2) see a 25 23-90 0040 | _... | +0°15
2 130 afecc tral® 25 23°90 0120 SUE EE Os
3 Jie 7) Ay tise da 25 23°80 0250 | .... | +0°05
4 Pee rcieese. | 10 25 23°15 0400 | ..-. | +0°00
5 - 130 eateerss lt ek) 25 23°76 0500 Beaton eo a) WL
6 130 He Lae 10 25 23°70 "1000 ---. | —0°05
7 BBO pe) lets oA, AO 25 23:75 C7: 0) ES aa 0:00
8 TSO treat 5 LO 25 24°20 mes 0200! +0°45
9 torso Pee ee 25 23°95 jee 0200) +0°20
10 130 an} 10 25 23°80 Ea "0400 | +0°05
ll 130 Sao hee $20 25 23°75 Bes 0400 0°00
12 130 leeeree oe patito 1 25 23°15 pe "0400 0°00
13 130 ee ere EO 25 23°15 TOC bos. on 0°00
14 | 130 pee 10 25 Za715 20000) |)" 2225 0°00
15/5) 1130 nS ea 25 23-75 | 3:0000 |} _.-. 0-00
16 130 1 bee 25 23°72 .--- /1:0000| —0°03
17 130 1 shape 25 23°74 ---- (2°0000; —0°01
18° | |" 130 hl hacia 25 23°72 ....- |3°0000| —0°03
19 130 2 peewee 25 23°70 .--. |0°5000} —0°05
20 130 3 ets 25 3°15 pve vil O' SHO0 0:00
Temperature about 80°.
21 | 145 ean 10 25 |; 28°90" | -0:5000 |... | +0°35
ee ido.) 10.:|' 10° | 26 ash A ibadO! kee be
23 500 pa kOD teh ke 25 | 23°75 ROOR Tr Boo Oro
24 500 Pusat [i ene 25 | 23°70 | 1°0000 | .... | —0°05
25 | 600 Pro as |eekO 25 } 2410" 2) OpO00 | Li.) + Orde

466 Gooch and Peters—Titration of Ovalie Acid.

From these results it appears that the presence of a suitable
amount of manganous salt—either the sulphate (exps. 4-7,
13-15, 22-24) or the chloride (exps. 10-12, 16—20)—is capable,
either in cold solution (exps. 1—20) or in hot solution (exps. 22-24)
of preventing the action of the permanganate upon the hydro-
chloric acid. It appears, also, that, for a given dilution and
strength of acid, less manganous salt is needed in the cold -
solution (exps. 4-7) than in the hot solutions (exps. 22—24).
Thus, in the hot solution, at a dilution of 145™ to 500
1 grm. of manganous sulphate must be present with 5°™* of
strong hydrochloric acid, with or without sulphuric acid ;
while in the cold solution 0:04 grm. of either the sulphate or
chloride is enough to secure adequate protective effect.
Experience showed, however, that 0°5 grm. or 1:0 grm. of the
manganous salt should be present in order to push the reaction
with reasonable speed in cold solutions.

Wagner* has made record of the increased evolution of
chlorine in oxidations of ferrous chloride by potassium per-
manganate in presence of various salts, of which barium
chloride was the most active. We have made some experi-
ments, therefore, to determine whether such action would
appear in the oxidation of oxalic acid in cold solutions con-
taining certain salts, and, if so, whether it would be preventable
by the presence of the manganous salt under our conditions
of working. From the results given in the accompanying
table, it is plain that the evolution of chlorine in cold solu-
tions is less in the presence of these salts than when hydro-
chloric acid is used without them, and that such evolution may
be entirely prevented (within the proportions of our work) by
the presence of 0°5 grm. to 1 grm. of manganous chloride.

Finally, it appears as the result of an investigation, that the
titration of oxalic acid by potassium permanganate in presence
of hydrochloric acid is ordinarily attended with some inaccu-
racy due to liberation of chlorine from the hydrochloric acid ;
that this tendency may be overcome by the presence of a man-
ganous salt—either the sulphate or chloride; that 1 grm. of
the manganous salt is enough to so affect the conditions of
equilibrium that titrations in moderate volumes (100™ to
500°™°*) and in presence of hydrochloric acid (5° to 15°™* of the
strong acid) may be conducted with safety and reasonable
rapidity, either with or without sulphuric acid, at the ordinary
atmospheric temperature.

* Loe. cit.

Gooch and Peters—Titration of Oxalic Acid. 467

VOLUME AT BEGINNING OF TITRATION = 140°™?,
Temperature = 20°-24° C.

Ammonium HCl. MnCls. /KMnO, Error.

oxalate. strongest 4H.0. | BaCl,. SrCle. CaCl. MgCle.) used.

. em?. cm?, grams. grams. grams. grams.) cm?. em?.
25 5 0°5 = an ae Pe. OCOD 0°00
25 5 nes 2 is Bs moa bag fa 8) + 1°40
25 5) aS 2 ee or -. |26°50 +0°45
25 5) Be us 2 a Le 26°53 +0°48
25 5 ce Be as 2 a) 26°36 +0°35
25 5) Re a ce bao He 2 |26°13 +0°08

|

25 5) O2B5 hi] 2 a andb ae 26°05 0°00
25 5 Oro ies eae! IGM Oles st 10705
25 3) Ones say hee: 9 2 pata 26°10 + 0°05
25 5 Orage yy ae as Dre et2 6°05 0°00.
25 10 AOE 2 Ee cites ie 26°10 +0°05
25 10 | Sy | aed ee 2 a ie 26°05 0°00
25 10 ie ey nd ahs 2 be 26°06 +0°01
25 10 1:0

be ae nit 2 {26°11 =+0°06

Am. Jour. Sor.--FourtH Srrizs, Vou. VII, No. 42.—Junz, 1899.
31

468 Scientific Intelligence.

SCIENTIFIC INTELLIGENCE.

I. CHEMISTRY AND PHYSICS.

1. On the Influence of Electric Oscillations on Vapors.—lt has
been pointed out by Kaurrman that when the vapors of various
chemical compounds are subjected to the influence of electric

oscillatory discharges the results obtained are: (1) the vapor |

becomes luminous, (2) it is crossed by spark discharges, generally
of a green color, or (3) it remains dark, at least until a very high
pressure is reached. More than fifty vapors were examined in
this way, and the results, while too few to permit of general
deductions, yet appear to indicate certain constitutive influences.
Thus, for example, the aromatic amines for the most part become
luminous and show quite similar appearances, the facility with
which luminosity occurs appearing to be partly dependent on
the basic character of the amine; since compounds which have
lost their basic character by substitution, such as nitraniline and
tribromaniline, no longer become luminous by the discharge.
Cyclic compounds, however, containing nitrogen, do not appear to
follow this rule. In the case of hydrocarbons, aliphatic com-
pounds and the more simple benzene compounds remain dark until
a very high pressure is reached, though most of the hydrocar-
bons containing two or more benzene rings readily become lumi-
nous. During luminosity, it was noticed, that in all cases the
vapor acted as a conductor.—Zeitschr. phys. Chem., xxvi, 719-
727, August, 1898. G, FoR:
2. On the Occlusion of Oxygen and Hydrogen by Platinum
black.—In experiments made by Monn, Ramsay and Surexps to
test the occlusion of oxygen and hydrogen in this way, the plati-
num black was saturated with hydrogen, heated at 184° under
reduced pressure, then placed in an ice calorimeter and allowed
to reabsorb the gas. It was found that 68-8 calories were evolved
for every gram of hydrogen occluded. In the opinion of the
authors the argument offered by Berthelot in support of the
existence of the compounds Pt,,H, and Pt,,H, are not justified by
the facts. Moreover they conclude that the difference between
palladium and platinum, in their behavior towards hydrogen, as
stated by Favre, is due simply to the presence of oxygen in the
platinum black. On attempting to remove the oxygen from

platinum black, in order to determine its heat of occlusion, it was -

noticed that reducing agents, while eliminating the oxygen, were
themselves occluded and retained obstinately. It was also
observed that hydrogen does not remove all the oxygen, but only
replaces that at first removed, both gases being simultaneously
occluded. On saturating platinum black with hydrogen and
exhausting it at 184°, it was found that, on allowing it to come in
contact with small quantities of oxygen, the addition of more
oxygen beyond a certain point, caused the pressure to increase.

4 La

7

»

4
j

4
|

Chemistry and Physics. 469

On adding oxygen, however, up to atmospheric pressure, a further
small quantity of the gas was absorbed and the heat simultane-
ously developed represented the true heat of occlusion of this:
quantity of oxygen. The results obtained were in close agree-
ment with others obtained indirectly, giving as a mean 11 calories
for every gram of oxygen occluded. This value agrees well with
that given by Thomsen for the heat of formation of platinous
hydroxide. Possibly, therefore, the two phenomena may be iden-
tical, the water required being present in platinum black dried in
a vacuum.

In a subsequent investigation, the authors studied the action of
palladium black in occluding gases. Prepared in the same way
as platinum black, it contains 1°65 per cent or 138 times its vol-
ume of oxygen. But this oxygen, unlike that in platiaum black,
cannot be removed by heating to dull redness in a vacuum, but
must be heated in a current of hydrogen. When heated in
oxygen the palladium black absorbs it, up to a red heat, yielding
a brownish black substance, which does not lose its oxygen on
heating to redness in a vacuum. Nearly 1000 volumes are
absorbed, an amount about three-fourths of that required to form
the oxide PdO. Over 1000 volumes of hydrogen are absorbed by
palladium black, though only 873 volumes are occluded, the rest
having formed water with the oxygen present. About 92 per
cent is removed by the air pump at the ordinary temperature, the
rest at 444°. The palladium sponge from this experiment occluded
852 volumes of hydrogen, of which 98 per cent was extracted ina
vacuum at ordinary temperatures. ‘The heat of occlusion for
hydrogen was found to be 46°4 calories for each gram of the gas ;
or 43°7 if the work of the atmosphere be allowed for. The heat
developed per gram of oxygen occluded, determined indirectly,
was found to be 11-2 calories, a value intermediate between that
found by Thomsen for the heat of formation of palladious and ©
palladic hydroxides; favoring the idea that this occlusion of oxy-
gen is a true oxidation. The atomic ratio of palladium to
hydrogen in the fully charged metal varies between 1°37 and 1°47;
agreeing closely with the formula Pd,H, as suggested by Dewar.
—Proc. Roy. Soc., \xii, 50-52, 290-293; J. Chem. Soc., lxxiv,
ii, 599, 600, December, 1898. G. F. B.

3. On certain Derivatives of Acetylene.—It has been observed
hy Erpmawn and Koruner that when acetylene is passed over
finely divided copper heated to 400°-500° it is decomposed into
hydrogen and carbon, the latter appearing in the graphitic condi-
tion. At temperatures below 250°, however, the copper unites
with the gas, forming a yellowish-brown compound which is not
explosive. This new substance may be more readily prepared by
heating finely divided cuprous oxide in a current of acetylene at
250°. It is very bulky; and has the composition C,,H,,Cu,.
When heated with zinc dust in excess it yields 20 per cent of an
oil having an odor like Caucasian naphtha and boiling between
190° and 250°. Well-defined acetylides of the alkali metals are

x

Er

?
=

—

» 2
di
ae
enn
f “
q i y
@

470 Scientific Intelligence.

not easy to prepare. Rubidium, though acted on by acetylene at
low temperatures, gives rise to no definite product. Zine and mer-
cury are scarcely acted on, and iron, like copper, acts catalytically,
inducing condensation to oily hydrocarbons. Acetylene pro-
duces no precipitate in solutions of thallium, lead, cadmium, iron,
nickel, cobalt, platinum, iridium, and rhodium. Gold chloride
gives a black precipitate, changing to metallic gold on warming.
Palladium chloride gives a brown precipitate soluble in ammonia.
Copper acetate yields a brownish-red and silver nitrate a white
precipitate. By saturating a hot solution of mercuric nitrate with
acetylene, mercurocarbide nitrate HgC : CHg. HgNO,. H,O sepa-
rates in small white crystals. It yields acetaldehyde on treatment
with dilute acids.—Zeitschr. anorg. Chem., xviii, 48-58, 1898.
G. F. B.

4. On Hutropic Series of the Calcium Group.—In 1896 Linck
proposed the term catameric eutropy or more briefly eutropy, for
those special cases of isomorphism in which the geometrical and
physical characters of members of a series increase or decrease
regularly with increase of the atomic mass of the varying ele-
ment; as is the case for example in the potassium, rubidium and
cesium compounds described by Tutton. Subsequently he called
attention to the fact that the quotients obtained by dividing the
crystal volumes of the different members of such a series by their
molecular volumes are in the ratio of simple natural numbers.

These relations have been recently tested by Eppier, who.
finds that they both hold good in the calcium-barium-strontium.
group, with one or two exceptions, as in the case of the three
anhydrous sulphates, anhydrite, celestite and barite. In his paper
he has brought together and compared numerous data concerning
the crystallized compounds of this series of elements, giving new
geometrical, optical and density determinations. In the case of

cubic crystals the volume is unity and the above mentioned quo-

tient 1s inversely proportional to the molecular volume. In the
cubic oxides, fluorides, chlorides and nitrates of calcium, stron-
tium and barium, the ratios are 11:15:19, 14: 17:20, 18:19: 20,
and 20:21:22 respectively; thus decreasing with the increase in
molecular mass. Although lead compounds are frequently iso-
morphous with those of calcium, strontium and barium, they do
not belong to the same eutropic series, the physical constants fall-
ing between those of calcium and strontium or before or after
those of the barium salt corresponding.—Zettschr. Kryst. Min.,
xxx, 118-175; J. Chem. Soc., 1xxiv, 560, December, 1898.
G. F. B.

5. The distribution of Vanadium.—It was shown by HassEt-
BERG in 1897 (Astrophys. Jour., vi), that the mineral rutile con-
tains vanadium in small but varying amount. He has now (ibid.,
ix, 143) continued his investigation to a considerable number of
meteoric stones, and concludes from this that vanadium is always
present in them, although it has not been detected in meteoric
irons. A recent paper by Hillebrand (this Journal, vi, 209) has

a ai a ee ed ot -

r= | s&s -ae

a t,'
a .au2 eo

Chemistry and Physics. 471

shown that both vanadium and molybdenum are widely dis-
- tributed in rocks, the former probably present in the silicates of
the amphibole or pyroxene group and in the biotites.

6. Determination of time of vibration of very high notes.—
F. MELDE reviews the various methods that have been suggested
and used for such determinations and divides them into two
groups: A, the subjective; B, the objective. Under A fall the
determinations carried out by the human ear, such as the
comparison of sounds, one of which is used as a standard.
Determinations of notes of 6144 vibrations have been made by
this method; for higher notes the method becomes inapplicable.
To the subjective method also belongs the difference-tone method,
which consists in measuring a third tone by the use of two pri-
mary tones. Notes of 8692 vibrations have thus been measured.
To the objective method belong the various graphical and photo-
graphic methods. In conclusion the author points out the applica-
bility of sensitive flames for the measurement of high notes.—
Wied. Ann., No. 4, 1899, pp. 781-793. aah

7. Heat conductivity of different kinds of glass—A great
number of determinations have been made by Focke and by
Paalhorn of the thermal conductivity of glass which differ greatly
in value. A. WINKELMANN endeavors to reconcile the results
and concludes that it cannot be decided which determinations are
the better. The discrepancies are in part due to the different
ranges of temperature employed. The author concludes that it is
desirable to investigate further the dependence of the conductivity
of glass upon temperature.— Wied. Ann., No. 4, 1899, pp. 794-802.

Eg

8. Absorption of Electric Waves by different substances.—The
recent experiments in wireless telegraphy have excited interest in
the question of the electrical opacity of non-metallic substances.
EK. Branty and G. Legon have undertaken a research upon this
question. A coherer together with a battery and a bell were
placed in a cavity in the substance which was to be investigated.
The substances employed were stones, blocks of cement, and
boxes of sand. The electric waves were excited by a Rhigi oscil-
lator placed outside the substance. It was found that the opacity
depends upon the nature of the substance. The transparency is
very great in the case of sand and the kind of building stone
employed. It was extremely feeble with Portland cement. The
opacity increases with the thickness. Humidity clearly increases
the opacity.— Comptes Rendus, April 14, 1899, pp. 879-882.

5 a

9. Radiations of Uranium.—H. BecquERet concludes fur-
ther study of these radiations by the following conclusions: The
radiation of radio-active bodies approaches approximately in
character to X-rays more nearly than to ordinary light. The
most singular fact connected with these radiations is the spon-
taneous emission of such radiations without known cause. If it
could be shown that this radiation occurs without the consump-

472 Scientific Intelligence.

tion of energy, one could compare the state of uranium to that
of a magnet. We can also compare the state of uranium to that ~
of phosphorescent bodies which seem to conserve their state of phos-
phorescence indefinitely. The photographic and phosphorescent
effect, however, of the uranium radiation indicates a consumption
of energy. This consumption is extremely feeble, and seems to
indicate that uranium can emit its unknown radiation for a long
period of years apparently without sensible diminution.— Comptes
Rendus, March 27, 1899, pp. 771-776. aL
10. Source of Uranium Radiations.—Prof. W. Crookes sug-
gests that uranium, thorium, polonium and radium have the power
of separating rapidly moving molecules from slowly moving ones
and appropriating a portion of the energy of the former. This
energy may be used in dissociation and maintaining radiations in
the ether. The air in the neighborhood of such substances would
be cooled. This cooling is so small as to escape detection, and
therefore the energy appears to be created out of nothing. The
author computes that air contained in a room.4X8X/7m. could
supply one h.p. for fifteen hours.— Comptes Rendus, 128, pp.
176-178, 1899. ae
11. Phosphorescence at low Temperatures.—A. and L. LuMImRE
have found that the phosphorescence of calcium and zine sulphides
-produced by exposure to an arc light disappears wholly at about
—50° C. When exposed to a magnesium light it disappeared
between —70° and —90°. When the exposure to a magnesium
light is made at —191° C. the temperature of liquid air and the
body is heated, phosphorescence begins to appear at —180° C. aud
increases with the heating. Réntgen rays act like ordinary light.—
Comptes Rendus, 129, pp. 549-552, 1899. 5 Oe
12. Recueil Données numériques publié par la Societe Prangaise
de Physique. Optique par H. Durer. Deuxiéme fascicule. Pro-
priétés optiques des solides; pp. 417-785, Paris, 1899 (Gauthier-
Villars).—This volume contains tables of physical constants in
the department of optics, prepared under the direction of the
Physical Society of France. ‘The opening tables give the refrac-
tive indices for different wave-lengths of certain remarkable sub-
stances, as calcite, quartz, fluorite, halite, and so on. Then
follows a table of refractive indices for certain important samples
of glass. After this we have an alphabetical list of inorganic
solids, largely mineral species, with their optical properties
arranged in tabular form; references are given to the original
publication where the facts were given. This portion covers
pages 466 to 632. A like table for organic solids follows, and
then other tables giving the influence of temperature on optical
properties, and allied subjects, close the volume. This enumera-
tion of its contents will show how much this valuable book of
reference contains.

Miscellaneous Intelligence. 473

Il. MiIscELLANEOUS SCIENTIFIC INTELLIGENCE.

1. Lethyologia Ohiensis by C. S. Rafinesque, a verbatim reprint;
by R. E. Catt, 8vo, 175 pp. Cleveland, 1899. (The Burrows
Brothers Co. )—Rafinesque’ S genera and species, whether based on
actually existing forms or Audubon’s cruel jokes, or in later life
evolved from his own diseased mind, have been a stumbling block
to the systematic zoologist for more than half a century. The
extreme rarity of many of his published works has increased the
difficulty of recognizing the forms he described and has doubtless
often forced subsequent authors to ignore him unwillingly. The
late G. W. Tryon many years ago republished Rafinesque’s con-
chological works, and now Dr. Call and his publishers give us
the work on American fishes in this handsome volume, issued in a
limited edition of 250 numbered copies. The volume contains a
brief biographical sketch of Rafinesque, an account of his ichthyo-

logical work, a verbatim reprint of the Ichthyologica Ohiensis (of

which only eight copies are now known to exist), a bibliography
of his papers on fishes, and a reproduction of a previously unpub-
lished letter from Rafinesque to Dr. Daniel Drake of Cincinnati,
containing a sketch of one of the fishes described. It seems
ungrateful to criticise a work of this kind for its omissions, but a
general index is much needed, the reprint of the original index of
the Icthyologica with references to the genera and English names
being almost useless for reference.

If the “ excentric naturalist ” could have foreseen the appear-
ance, before the end of the century, of this luxurious volume
devoted to a reprint of the work to which he gave himself with
such prodigality of unrewarded zeal, he might have forgiven and
forgotten the neglect and scorn of his cotemporaries and retained
his reason through the unhappy later years of his life. s.r. 5s.

2. Elementary Physiology; by Brnsamixy Moore, M.A.
pp. 295. New York, 1899. (Longmans, Green & Co.)—This
elementary text-book differs in several respects from many other
books of its class. First, a relatively large portion of the text is
devoted to the description of the structure of the animal body.
Since the functions of the organism are so closely connected with
its morphological characters, a somewhat comprehensive survey
of the latter becomes highly essential to a well-conducted course
in physiology for beginners. The usefulness of the present text-
book is greatly increased by an unusually large number of well-
selected illustrations. Second, in the consideration of the various
departments of physiology due appreciation of the relative im-
portance of the phenomena discussed has usually been shown.
Too many of the elementary text-books are unsatisfactory in this
respect: they either unduly emphasize certain actions of the body,
or omit entirely chapters which do not permit of easy treatment.
Professor Moore’s little book certainly does not err in the latter
respect, and it may even be questioned whether occasional topics

474 Scientific Intelligence.

have not been introduced with too great detail for the beginner.
Third, the reader is not allowed to forget that physiology is an ~
experimental science and should properly be taught as such ;
accordingly, the general method of the book tends to encourage
observation and experiment. A series of carefully defined exer-
cises forms a useful appendix. L. B. M.

3. Hxperimental Morphology, Part II; by C. B. Davenport.
New York, 1899.—In this second installment of his valuable
treatise Dr. Davenport considers the effects of chemical and
physical agents on growth. Plants lend themselves so readily to
observations respecting growth, that they have received from Dr.
Davenport a degree of attention in this work which compels the
student of plant physiology to accept his aid. The literature of
the whole province has been carefully examined, and few omis-
sions can be noted in the references. On the whole, the editorial
work has been wonderfully well done, and the errors are so few
that they need not be mentioned here; they do not appear likely
to mislead seriously any students who make use of the vast
amount of material which has been gathered patiently, and for
the most part has been well arranged. The botanical phases of
experimental morphology are at present attracting a large num-
ber of enthusiastic students who are moved to consider questions
in a broad way, and are stimulated by just such books as Dr.
Davenport’s, to ask some of these questions, at first hand, of
nature. Gi E.1Gs

4. Recherche, Captage et Aménagement des Sources Thermo-
Mineérales. Origine des eaux Thermo-Minérales; Géologie; Pro-
priétés physiques et chimiques. By L. De Launay, pp. x, 635,
large 8vo. Paris, 1899 (Librairie Polytechnique Baudry et Cie.,
éditeurs).—M. De Launay has added another to the series of
highly valuable practical works prepared by him. The previous
volumes on Mineral Deposits, on the Gold Mines of Transvaal, and
the Diamonds of South Africa, have been noticed in earlier num-
bers of this Journal. The present volume discusses thermal
springs in different parts of the world. The treatment is thor-
ough and exhaustive. The origin of the springs and their rela-
tion to geological and topographical features is first taken up.
Then comes a discussion of their chemical composition with the
salts and gas in solution, accompanied by numerous analytical
tables. Following these are data in regard to the temperature at
different springs, ranging, for example, from 81°:5 at Chaudes-
Aigues (Cantal) to 27° at Uriage (Isere). The latter half of the
work goes more minutely into the account of the individual locali-
ties in different parts of the world, giving a large amount of very
interesting information about a multitude of places. The work
closes with details in regard to the distribution of the hot water
to the different establishments, with numerous excellent illustra-
tions.

OBITUARY.

Dr. Franz Rirrer von Hauer, the distinguished Austrian
geologist, died March 20th, at the age of seventy-seven years.

IPE X: TO VOR ViIt*

A

- Absorption of light in a magnetic
field, Rhigi, 239.

Academy, National, meeting at Wash-|
ington, 401.

Africa, origin of culture in, Frobe-
nius, 80.

Air, stratified brush discharges in,

Toepler, 67.

Alabama, iron-making in, Phillips,
398.

Ames, J. S., Prismatic and diffraction.
spectra, 69.

Arizona, Governor’s report, 169.
Austin, M., ammonium-magnesium
phosphate, 187.

B

Fruits, 78.

Barnes, C. R., Plant Life, 244.

Barus, C., thermodynamic relations
of hydrated glass, 1.

Beach, F. E., General Physics, 314.

Beecher, C. E., notice of Othniel
Charles Marsh, 403.

Bolton, H. C., Bibliography of Chem-
istry, 1492-1897 » d22.

Botany, elements of, Van Tieghem,
77

Botany—

Bush-fruits, Card, 78.

Caulerpze, monograph of the, Bosse,
400.

Cyperacez, Studies of, Holm, 5,
171, 435.

Delesseria, Agardh, 401.

Flora of the West Indies, Urban,
244.

Magnolias, formation of pollen-
grains, Guignard, 77.

Marine alge of Greenland, Rosen-
vinge, 400.

_ Botany—
Morphology, Part II, Davenport,
474.
Native Fruits, evolution of, Bailey,
78.
Pfianzenfamilien, die Natiirlichen,
Engler, 169.
Plant Life, Barnes, 244.
‘Phenols, effect on living plants,
True, 76.
Rhodora, 245.
Root-pressure, cause of, Leavitt,
301.
Brazil, argillaceous rocks and quartz
veins in, Derby, 343.
Brush discharges in air, Toepler, 67.

C

Cajori, F., History of Physics, 394.
Bailey, L. H., Evolution of our Native |

Call, R. E., Rafinesque’s Ichthyologica
Ohiensis, 473.

: - s ~
Canada, geol. survey, vol. ix, 71.

Canadian Paleontology, Whiteaves,
ch.
Card, F. W., Bush-fruits, 78.
Chemical Analysis, Manual, Newth,
67.
CHEMISTRY—
Acetylene, certain derivatives of,
Erdman and Kothner, 469.
Aétherion, Crookes, 64.
Alkali nitrates, preparation, Divers,
dll.
Ammonium-magnesium phosphate,
Gooch and Austin, 187.
Beryllium, electrolytic preparation,
Lebeau, 155.
Boric acid, estimation of, Gooch
and Jones, 34; Jones, 147.
Calcium, metallic, Moissan, 393.
group, eutropic series, Eppler,
471.
Cuprous chloride and acetylene,
Chavastelon, 237.

* This Index cofitains the general heads, BOTANY, CHEMISTRY (incl. chem. physics), GEOLOGY,

MINERALS, OBITUARY,
tioned,

20CKs, and under each the titles of Articles referring thereto are men-

476

CHEMISTRY—
_ Gases, molecular masses, Berthelot,
154,
Graphitic acid, Staudenmaier, 65.
Helium, homogeneity of, Ramsay
and Travers, 510.
Hydrochloric acid in titrations by
sodium thiosulphate, Norton, 287.
Todine, spectra of, Konen, 67.

Tron chlorides in analysis, volatili-
zation, Gooch and Havens, 370.
Lithium and calcium with ammo-

nium, Moissan, 69.
Metals, preparation by means of
aluminum, Goldschmidt, 104.
Methane and air, explosion by elec-
tric current, Couriot and Meunier,
236.

Neodymium, Boudouard, 197.

Nickel, extraction by the Mond
process, Roberts-Austen, 64.

Oxalic acid, titration, Gooch and
Peters, 461.

Oxygen and hydrogen, occlusion,
Mond, Ramsay and Shields, 468.

Ozone, properties, Ladenburg, 310.

Palladium,
of, Zelinsky, 393.

Petroleum, composition of Ameri- |

ean, Young, 311.
Platinum and gold, solution in elec-
trolytes, Margueles, 236.
Silver, colloidal, Lottermoser and
Von Meyer, 156.
peroxide and peroxynitrate,
Mulder, 156.
Strychnine, new bases from, Tafel,
394.
Vanadium, distribution, 470.
Chemistry, Bibliography, 1492- 1897,
Bolton, 822.
Industrial, Thorp, 157.
Lessons in physical, Van’t Hoff,
157.
Clarke, F. W., hydromica from New
Jersey, 365; composition of roscoe-
. lite, 451.
Clements, J. M., contact metamorph-
ism, 81.
Coherer, use in measuring electric
waves, Behrendsen, 158.

Cycads, American fossil, Wieland,
219, 305, 385.
D
Darton, N. H., hydromica from New

Jersey, BOO. <i

Davenport, C. B., Morphology, Part
II, 474.

Davis, W. M., Physical Geography, 248.

reductions in presence |

INDEX.

Davison, J. M., platinum and iridium
in meteoric iron, 4,

se Hai Sources thermo-minérales,

74,

Derby, O. A., argillaceous rocks and
quartz veins in Brazil, 348. ~

Dielectrics, polarization and hyste-
resis, Schaufelberger, 312.

Diller, J. S., origin of Paleotrochis,
Boe

Dufet, Optical physical data, 472.

Duhem, P., Thermodynamics, 68.

x

Harth, age of, Kelvin, 160.

Earth Sculpture, Geikie, 72.

Eastman, C. R., Devonian Ptycto-
dontidea, 314.

Electric oscillations, measurement of
slow, Konig, 3895; influence on
vapors, Kauitman, 468.

spectrum, dispersion in, Marx, 68.
waves, absorption of, Branly and
Lebon, 471.
new method for showing, Neug-
schwender, 812.
use of the coherer in measuring,
Behrendsen, 158.
along wires, Sommerfield, 311 ;
Coolidge, 396.

Electricity, conduction through gases

by charged ions, Thomson, 812.
carried by the ions produced by
Réntgen rays, Thomson, 158.

Electrolytic interrupter, Swinton, 395.

Endothermic gases, experiments with,
Mixter, 323.

Energy, luminous
Berthelot, 309.

and chemical,

Evolution, organic, Fairhurst, 80.

F

Fairchild, H. L., glacial lakes in Cen-
tral New York, 249.

Field Columbian Museum report, 248.

Florida, Tertiary fauna of, Dall, 71.

Foote, i. We, composition of tour-
maline, 97.
Funafuti, Sollas, 317.
G

Gases in a high vacuum, absorption,
Hutchins, 61.
evolved on heating mineral sub-
stances, Travers, 244.
endothermic, Mixter, 323;
plosion of, Mixter, 327.
Geikie, J., Earth Sculpture, 72.

ex-

ag
_

INDEX.

Geographical congress, International,
6.

Geological congress,
316
GEOLOGICAL REPORTS AND SURVEYS—
Alabama, Phillips, 398.
Canada, vol. ix, 1896, 71.
Iowa, 1897, 168.
Maryland, Clark, 69.
South Dakota, 316.
United States, 166.
West Virginia, White, 398, 399.
GEOLOGY.—
Camden chert of Tennessee, Safford
and Schuchert, 429.
Carboniferous invertebrates, index
of North American, Weller, 70.
Crater Lake, Oregon, Diller, 316.
Cretaceous and Tertiary plants of
North America, Knowlton, 168.
Cycads, American fossil,
219, 305, 383.
Devonian fishes, Eastman, 314.
Earth movement in the Great Lakes
region, Gilbert, 239.
Kocene carnivore, Marsh, 397.

International,

Franklin white limestone. New Jer- |

sey, Wolff and Brooks, 397.

Geological sketch of San Clemente
Island, Smith, 315.

Geology of the Edwards plateau,
ete., Hill and Vaughan, 315.

Glacial Lakes in Central New York,
Fairchild, 249.

Jurassic Dinosaurs,
Marsh, 229.

Loess, origin discussed, Sardeson,
58.

footprints,

Lower Cretaceous Grypheas of |
Texas, Hill and Vaughan, 70.

Lytoceras and Phylloceras, develop- |

ment, Smith, 398.
Metamorphism, contact, Clements,
81

Paleotrochis in Mexico, Williams,
335 ; origin of, Diller, 337.
Saurocephalus, Hay, 299.

Tertiary fauna of Florida, Dall, 71. |

Trenton rocks at Ungava, Whit-
eaves, 433.
Wind deposits, mechanical compo-
sition, Udden, 74.
Georgia, gold deposits, 168.
Gesteinslehre, Rosenbusch, 73.
Gilbert, G. K., earth movement in
the Great Lakes region, 239.
Glass, thermodynamic relations of hy-
drated, Barus, 1.
Gold, metallic, aqueous solutions,
Zsigmondy, 236,

Wieland, |

Gooch, F. A., estimation of boric
acid, 34; ammonium-magnesium

phosphate, 187; volatilization of
iron chlorides in analysis, 370;
titration of oxalic acid, 461.

Gould, Benjamin Apthorp, fund, 320.

Gravitation in gaseous nebule,
Nipher, 459.

Greenland, marine alge of, Rosen-
vinge, 400.

H

Hall phenomenon, Poincaré, 312.
Harper’s Scientific memoirs, 69, 402.
Hastings, C. S., telescope objective,
267; General Physics, 314.
Havens, F. S., volatilization of iron
chlorides in analysis, 370.
| Hay, O. P., unknown Bee of Sau-
rocephalus, 299.
_Heat conductivity of Ese Winkel-
mann, 471.
long waves of, separated by quartz
prisms, Rubens and Aschkinass,312.
/ | Hill, R. T., Lower Cretaceous Gry-
pheeas of Texas, 70.
'Hillebrand, W. F., mineralogical
| notes, 51 ; analyses of biotites and
| amphiboles, 294; roscoelite, 451.
_ Hobbs, W... H. , goldschmiadtite, 307.
| Hoffmann, G. C., polycrase in Can-
ada, 243,
_Holm, T., Cyperacez, 5, 171, 435.
Holman, 8. W., Matter, Energy, Force
| and Work, 237.
| Hudson Bay region,
| rell, 72.
Hutchins, C. C., absorption of gases
| ina high vacuum, 61.

geol. report, Tyr-

I

Iowa geological survey, 168.

J

Jones, L. C., estimation of boric acid,
34, 147.

Judd, J. W., petrology of Rockall
Island, 241.

K

Kaolins and fire clays of Europe, Ries,
243.

Kunz,’ G. ¥F.,
Carolina, 242,

native silver in No.

=

Le

ee lle

oe
.

|
»
*
6 '
ey |
4
|
f ;
an
1 aes
4{
a {
| ‘
my! ahh
i)
1
\
i call hi
'
40
q
|
oP
§
|
}
\

. > s = r

478 | INDEX.

L

Leavitt, R. G., cause of root-pressure,
381.

Lightning, spectrum of, 68.

Liquid air, density and composition,
Ladenburg and Kriigel, 234.

Luquer, Minerals in rock sections, 319.

M

Mammals, origin of, Osborn, 92.
Marsh, O. C., footprints of Jurassic

Dinosaurs, 229; note on a Bridger

Kocene carnivore, 397; obituary no-

tice by C. HE. Beecher, 403; biblio-

eraphy of, 420.

Maryland Geol. Survey, 69.
Matter, Energy, Force and Work,

Holman, 287.

Means, T. H., determination of solu-
ble mineral matter in a soil, 264.
Membranes, semi-permeable, Mijers,

280.

Meteoric iron, containing platinum

and iridium, Davison, 4.

Meteorite, Ness Co., Kansas, Ward,

239.

Meteorites, gases yielded by, Travers,

244. .

Micas, crystal symmetry of, 199.
Minerals in rock sections, Luquer, 319.
MINERALS—

Amphibole, 297.

Bastnisite, Colorado, 51. Batavite,
76. Biotite, 202, 294.

Cedarite, 76. Chromite, North Car-
olina, 281. Corundum in India,
and in Canada, 318. Covellite,
Montana, 56.

Enargite, Montana, 56.

Gold of Georgia, 168. Goldschmidt-
ite, 307. Griinlingite, 76.

Hydromica, New Jersey, 365.

Jeffersonite, 55.

Ktypeite, 320.

Lagoriolite, 519. Lepidolite, 202.

Mitchellite, 286. Mossite, 75. Mus-
covite, 208.

Phlogopite, 201. Planoferrite, ‘76.
Polycrase, Canada, 243. Powel-
lite crystals, 367. Prosopite, Utah,
53

Roscoelite, 451.

Senaite, 75. Silver, No. Uarolina,
242. Sulphate, fibrous, Mon-
tana, 57.

Thalenite, 820. Tourmaline, 97.
Tysonite, 51.

Valléite, 75.

Zinnwaldite, 202.

Mixter, W. G., experiments with
endothermic gases, 823; partial
non-explosive combination of ex-
Le gases and gaseous mixtures,

Mont Blanc, geology and petrog-
raphy, Dupare and Mrazec, 242.

Moore, B., Physiology, 473.

N

New Guinea, etc., Zoological studies,
Willey, 79, 322.

Newth, G. S., Manual of Chemical
Analysis, 67.

Nipher, F. E., gravitation in gaseous
nebule, 459. . ‘

Norton, J. T., Jr., hydrochloric acid
in titrations by sodium thiosul-
phate, 287.

Norway, Crustacea of, Sars, 79.

0
OBITUARY—
Hauer, Franz Ritter von, 474.
_ Marsh, O. C., 403.
Peck, L. W., 248.
Wiedemann, G., 402.
Optical instruments of R. Fuess,
Leiss, 396.
Osborn, H. F., Origin of mammals,
92.
Ostwald’s Klassiker der exacten Wis-
senschaften, 170.

P

Palache, C., powellite, 367.

Patagonia, mollusks from, Pilsbry,
126.

Penfield, S. L., composition of tour-

_ maline, 97.

Peters, C. A., titration of oxalic acid,
461. ;

Phosphorescence at low temperatures,
Lumiére, 472.

Physical Geography, Davis, 248.

Physics, General, Hastings and Beach,
314.

History of, Cajori, 394.

Physiology, Moore, 473. —

Pilsbry, H. A., mollusks from Pata-
gonia, 126.

Pirsson, L. V., phenocrysts of intru-
sive igneous rocks, 271.

Pratt, J. H., chromite, origin, etc. of,
281.

INDEX.

BR
Rafinesque, Icthyologia Ohiensis, 473. _
Regnault’s calorie, Starkweather, 13.
Rivers of North America, Russell, 72.
Rock specimens distributed by the
U.S. Geol. Survey, Diller, 74.
Rockall Island, petrology, Judd, 241.
Rocks—
Argillaceous, with quartz veins in
Brazil, Derby, 348.
Clay slates, phyllites, etc., contact
metamorphism of, Clements, 81. |
Euctolite, 399.
Granitic breccias of Grizzly Peak, |
Colo., Stone, 184. /
Hatherlite, 318. /
Igneous rocks containing Paleotro-_
chis, Diller, 337. |
Pre-Cambrian of Wisconsin, |
Weidman, 398. |
Mont Blane petrography, Dupare
and Mrazec, 242. °
Norites, gabbros, etc., of Trans-
vaal, Leo Henderson, 317.
Olivine-melilite-leucite rock, Saba- |
tini, 399.
Phenocrysts of intrusive igneous
rocks, Pirsson, 271.
Pilandite, 318. |
Rockallite, 241.
Venanzite, 399. |
Rollins, W., regenerating vacuum |
tubes, 159. |
Roéntgen-rays, absorption by air, 396. |
and color blindness, Dorn, 159.
ions produced by, Thomson, 158.
vacuum tubes for, Rollins, 159.
Rosenbusch, H., Gesteinslehre, 73.
Russell, I. C., Rivers of North Amer-

|

ica, 72.

S |

Safford, J. M., Camden chert of Ten-|
nessee, 429. |

San Clemente Island, geological
sketch, Smith, 315.

Sardeson, F. W., What is the Loess ? |
38

58. /
Schuchert, C., Camden chert of Ten- |
nessee, 429.
Smithsonian Institution, reports, 80, |
246, 321, 402.
Soils, analysis of, Means, 264.
South Dakota geological survey, 316.
Spectra, prismatic and diffraction,
Ames, 69.
Spectrum of lightning, Toepler, 68.
Springs, thermo-mineral, De Launay, |
474.

Trouessart,

479

Starkweather, G. P., Regnault’s cal
orie, and _ specific volumes of
steam, 18; thermo-dynamic rela-
tions for steam, 129.

Stars, Washington catalogue of, 321.

Steam, specific volumes of, Stark-
weather, 13 ; thermodynamic rela-
tions for, Starkweather, 129.

Stokes, H. N., analyses of biotites
and amphiboles, 294.

Stone, G. H., granitic breccias of
Grizzly Peak, Colo., 184.

Surface tension and density of aque-
ous solutions, etc., MacGregor, 318.

T

Telescope objective, Hastings, 267.

Temperature, low, experiments, Hem-
pel, 392.

Tennessee, Camden chert of, Safford
and Schuchert, 429.

Tennessee, university of, record of
scientific engineering, 170.

Texas, geology of, Hill and Vaughan,

Thermodynamic relations of hydrated
glass, Barus, 1.

relations for steam, Starkweather,

13, 129.

Thermodynamics, Duhem, 68.

Tho F. H., Industrial chemistry,
1957.

Transvaal norites, gabbros, and py-
roxenites, Leo Henderson, 317.

i Catalogus mammalium,
79.

Turner, H. W., rock-forming biotites
a amphiboles, 294; roscoelite,

5d.

U

Ungava, Trenton rocks at, Whiteaves,
433

United States geological survey, re-
cent publications, 166.

Uranium radiation, Rutherford, 238 ;
Becquerel, 471 ; source of, Crookes,
472.

V

Vacuum tubes, regenerating, Rollins,

159.

cel

woes

-

= me 2s eee 8

7 1 = , . ; a = =, = = . SS - —- wee - = 5 ae). — Ss - =. => = - eS — Ss ——— =
2 ate mo Be Ee mm Fo tl QE EEE EEE <a = ~— (= oe ae eae ES SSeS Se SS SET

480 INDEX.

Vibration of very high notes, time of,
Welde, 471.
Voltameter, silver, Kahle, 239.

Ww

Walker, T. L., crystal symmetry of
the micas, 199.

Ward, H. L., new Kansas meteorite,
"233.

Water, absence of coloration in,
Spring, 318.

Whiteaves, J. F.,
Ungava, 433. .

West Indies, flora, Urban, 244.

West Virginia geol. survey, 398, 399.

Wieland, G. R., American fossil
eycads, 219, 305, 388.

Williams, H. S. Paleotrochis i in Mex-
ico, 335.

Trenton rocks at

Xx

X-rays, see Rontgen rays.

t

r

Z

ZOOLOGY—

Actinians, new, Verrill, 41, 143,
205, 375.

- Asterias pallida, metamorphosis,

Goto, 78.
Crustacea of N orway, Sars, 79.
Fishes of North and Middle Amer-
ica, Jordan and Evermann, 79,
169.

Ichthyologica Ohiensis by Rafin- — :

esque, Call, 478.
Lepidoptera Phalene in the British
Museum, Hampson, 246. ;
Mammals, ‘catalogue of, Trouessart,
OES 2m: eek
Mollusca of the Chicago area, Baker,

79. :
Mollusks from Patagonia, Pilsbry,
126.
Zoological studies in New Chae
ete., Willey, 79, 322.

_ Am. Jour. Sci., Vol. VII, 1899. Plate |.

mE SP ERC SE EE ETE Pa

vi Am. Jour. Sci., Vol. VII, 1899. Plate II,

CYCADEOIDEA INGENS, Ward (type).

Am. Jour. Sci., Vol. Vil, 1899. Plate Ill.

CYCADEOIDEA INGENS ; type specimen.

.
Longitudinal section through male flower.

IV.

Plate

transverse sections through male flower.

Am. Jour. Sci., Vol. Vil, 1899.
CYCADEOIDEA INGENS ;

ay TS AS. AS

Am. Jour. Sci., Vol. VII, 1899. Plate V.

‘SUAVSONIG: OISSVUNOF HO SININALOO

—— UT Le ee

er <

8 s
1
{

ve q
oP iss | ay ae

is sheer

Se ee

a,
at), a

Am. Jour. Sci., Vol. Vil, 1899.

Plate VI.

oe re, = o
ees!) 87mm OO 240

{o)

ousn meray.

¢

y Ps)
oe

i)
Ly
CJaneuns?

ore,

a

a

a
a
a
laa
e

2

SHORELINES AND OUTLETS
OF GLACIAL LAKES

IN CENTRAL NEW YORK

——

77° PENNSYLVANIA

Am. Jour. Sci., Vol. VII, 1899.

oo
is
5

SSS ee
SS

!

$
eacet

I
J
'
I
-
d

CYCADEOIDEA INGENS, Ward (type); leaf structure

Plate VII.

Am. Jour. Sci., Vol. VII, 1899. Plate VIII.
1 2

Plate IX.

Vil, 1899.

Sci., Vol:

Am. Jour.

= eer: « — <

ne ae

Am. Jour Sci., Vol. VI!, 1899. Plate X.

We have just aecsivad a good-sized lot of specimens of
the most-beautiful Orpiment we have ever secured. The
specimens are all of small cabinet sizes, rarely exceeding
2x2 inches, and averaging about 1}x2. ‘The whole speci-
men is pure Orpiment and with the cleavage so perfectly ©

-dereloped that it can be split up into thin cleavage plates
with almost as much ease as Selenite, the thinner plates
being transparent and flexible. The lustre of these
cleavages is intensely brilliant and their color being of
the brightest yellow, our present specimens undoubtedly
will ke admired by collectors as the most attractive
specimens of this mineral ever seen. Prices? Cheap!

‘100. to 50c., mostly 20c. to 35c.; postage 2c. to 6c. extra.

Finest Amber Barites Ever Imported.

About the middle of June we expect to place on sale a new find of the English os
_ Amber Barites claimed to far excel any previously found. Also a few of the
| new and attractive “ Jacktrees’” Calcites and a good lot of Stank and Bigrigg |
Gu _ Mine Calcites, F luorites, ete. $ ayy .

a : Matchless Labradorite Specimens.

4 g
Pate One of our most prominent collectors after purchasing over $11 worth of pol-
ished Labradorite specimens from us said that we were not advertising them as
__ enthusiastically as their exceeding fine quality deserved. Possibly this has been
' becavise we had only rather larger specimens than most collectors want. Now
we have had a large number of “small and medium size specimens finished up and
> ean surely please everyone The prices will be about 10 cents per square inch,

believe, cheaper than they have hitherto been offered, while the quality is
Per Reedinely fine! Blue is, of course, the predominating color, but green, yel-
_low, brown, purple, maroon, and occasionally even red colors are ‘frequently seen
in abundance, Probably the most attractive of all are those which show long le
‘narrow bands pier matey dull and of brilliant colors. EAS

ba

Other Recent Accessions.

Cylindrite i in choice, large specimens, $1.00 to $3.50.

2 _ Franckeite, the other new Bolivian tin mineral, also in large, fine pieces, FRY
«$100 to $3.50. eee.”
ti From Montana. Lustrous Vesuvianite crystals in blue calcite, 25c.to 75e.; =
_-—~—s« groups of gray and brownish red Vanadinite crystals, 25c. to $1. 00. oS.
From Colorado. Crystallized Enargite, good eroups, 25¢. to $3.00; crys- —
tallized Covellite, 25c. to $1.00. pare

From Utah. A few choice crystallized Clinoclasite, Tyrolite,Scorodite,

Brochantite, Olivenite, etc.

Paragonite, Bohemia, transparent twin crystals, with deep reentrant angles, a
25c. to 75c.; choice, simple prisms, 1Uc. to 35c., Rog
A large cousignment of German minerals now in the Custom House will soon Part

be ready. A

: GEO. L. ENGLISH & CoO., Mineralogists, ies
Pig REMOVED TO rT ek 4 : dey

812 and 814 Greenwich Street (S. W. Corner of Jane Street), New York City.

i
j
i

XLVUL —Studies 1 in the Cypeebens ies Hona.- a

CONTENTS.

J. F, Wuarrraves 22 on ee Oe et Ur

y.

XLIX.—Roscoelite ; by W. F. Hittesranp and ‘Hy. Wa
TURNER, with a note on its chemical constitution by F.

LI.—Titration of Oxalic Acid soe Potaaa Permmansiaae a
in presence of Hydrochloric Acid; by F. A. Goocn a

C. A. PETERS

,

Radiations of Uranium, H. BEcQuEREL,, 471.—-Source of Tiraasaeee Radia

W. Crookes: Phosphorescence at Low Temperatures, A. and L. Lume
Recueil Données numériques publié par la Société Francaise de Physi
DUFET, 472.

Miscellaneous Scientific Patcliacheee Ueiiy liens Ohiensis by ©. 8. Rennes 16
E. Catz: Elementary Physiology, B. Moors, 473. Re More

Taaetih Anerdlee. L. De nie. 474,

Obituary—Dr. FRANZ RITTER VON Haver, 474.

INDEX TO VOLUME VII, 475.

gir

== :

eri

[23 =e
— :

- Semmes Se St ee

Wy i te Teen 6a we P-L) AR OMS

| \ vYS Yyyers 4 he ek BUR | qt
Ty: \4 "ond 9 Ob) <A. Se ol Paes ly hye tt ping rite ' ui dp Henryl ga 4a

ure wid eesti
tae gad? Win ieeenres Ne ypoudypypooranr 7 ee he nn ther! : Le niall

a

+ Prcgn OO
| . Tle |
nes at Nan of) “& . e ali | Laas ee - 1 St intel lair jel F ere 49 ? \ tig
"mA wv. Rel) . ~Sy oo een or re : oS ¢ f
: vas. ‘ ‘ ms . T eqn \ dis ha A | a ‘| ™ 44 A.
a NG Tet TTT PH ng IN-»y: E aya Na alegt Boley vd dB TTT Pt vd ye le PBatis
are we Moria Dee a ea es “0 nN \ . ie yaek 4 ws abet os THM we bay
Py mene — Sanu 5 ee a Me ae’ pu |
fer. at yr y teal AR inn ce aa! ter ase. | matt ee
A: v ve:
AA. A way ; ry Pp j Wy
ea aN ¥ | whis® I 1 Ts ty 0 y) wt Ne - Sh, Ab
All ys [-qi-ab ., 2. gh. y Athy! Seth Hf PHAN, wart na asl Vou Nines li,
alee ra New “oe :
ina MUS tiga re vy AP Ay sesenl ph ras peer Raa
tae a ag *y,> an aia y, tt oe wee BBO + season |

o (a

ax weiss Rul \ nn ee al,
if fis, ide Wuyi t ie pede ey yg Seana sensed Si)
pve PARa 27 An Ee wie tect MReae 9 yas 0 rye has

<aré [eee : ren Bea a abm ae

«tics Oe r ya ,
Tees Oder tie.
el aa? , » AAO Ae aee dno nisi ‘na i

i, Oe OR 2 ‘ab O44 JS « % ° §

\
+o. ee Nok Soult aaa? acocaanavauyy'

Nae

Pig Oe ~— OP 10% Bp wth oy Bee
{ al , On or § vay ® * a ai “art N
er TT Ma Te AEs IAAT” by

sa enol,

i Ati Tht aaah AAS MLL AA. ry
a wih: ~ PUY se p: Te trey Lt ie
TY treet iy tt] w TIN Tt
AAO. Ras ~y Ath > pl\p Avie ; nal? \ -
} | \ itty Vy a uv) so jain witsp ests pog! i oe bi \ 4454.
aa iha’ oo} ub
ed Na ‘hs pApeiag a

SMITHSONIAN INSTITUTION LIBRARIES
WIAA
3 9088 01298 5586