print the first line of your file

samtools view test.bam | sed -n -e 1p
D3VDZHS1:98:C1URWACXX:1:1316:17052:8149    73    ctg9    1832    46 ...

print line 100000 of the file

samtools view test.bam | sed -n -e 100000p
D3VDZHS1:98:C1URWACXX:1:2104:15290:4476    185    ctg4    116422    45 ...

print all between

samtools view test.bam | sed -n '/D3VDZHS1:98:C1URWACXX:1:1316:17052:8149/,/D3VDZHS1:98:C1URWACXX:1:2104:15290:4476/p' >part1.sam

voila

wc -l part1.sam
100000 part1.sam

edit: it can be done much easier with split:

samtools view test.bam | split -d -l 100000 - part1_

but as Pierre suggesting (thank you!), you'll need to add headers later to make resulting sam files valid.

Just for fun, I just wrote a java program for this task. See https://github.com/lindenb/jvarkit/wiki/Biostar90204

Example:

$ java -jar dist/biostar90204.jar -m bam.manifest -n 3 -a 5 samtools-0.1.18/examples/toy.sam

$ cat bam.manifest
_splitbam.00001.bam    1    3
_splitbam.00002.bam    4    6
_splitbam.00003.bam    7    9
_splitbam.00004.bam    10    12

$ samtools-0.1.18/samtools view -h _splitbam.00003.bam 
@HD    VN:1.4    SO:unsorted
@SQ    SN:ref    LN:45
@SQ    SN:ref2    LN:40
@PG    ID:0    PN:com.github.lindenb.jvarkit.tools.biostar.Biostar90204    VN:7e17f8bd273cf081d4415bc4f579cd34e2c681d1    CL:-m bam.manifest -n 3 -a 5 samtools-0.1.18/examples/toy.sam
@CO    SPLIT:3
@CO    SPLIT:Starting from Read7
x1    0    ref2    1    30    20M    *    0    0    AGGTTTTATAAAACAAATAA    ????????????????????
x2    0    ref2    2    30    21M    *    0    0    GGTTTTATAAAACAAATAATT    ?????????????????????
x3    0    ref2    6    30    9M4I13M    *    0    0    TTATAAAACAAATAATTAAGTCTACA    ??????????????????????????